It is impossible to hit a winner [with the ITST Mod]

Everything about your online tennis matches !

Re: It is impossible to hit a winner [with the ITST Mod]

Postby rharm2 » 12 May 2020, 11:55

Hi Manu

In the original question the asker asked about the counter shot that falls deep into the baseline from a defensive position, is there any way we could see that reworked in a more realistic approach as to get a ball that deep from behind the baseline at the backboard is an extremely difficult task and in te its a regular occurence.
rharm2
crazy of the little yellow ball
crazy of the little yellow ball
 
Messages: 196
Gaming Since: 25 Feb 2020, 00:32
Location: United Kingdom

Re: It is impossible to hit a winner [with the ITST Mod]

Postby pidzi » 12 May 2020, 12:44

Hey good day,

Just finished test you suggested with Nadal, but first some responses to your observations/analysis.

the main goal was to do winners from the baseline, so nothing to "beat" here, and the secondary goal was to have a rally length as low as possible, and there you didn't beat the Chinse guy ; even if you remove the aces & double faults, he's at 5.4, while you got 6.7 when removing aces & double faults.


Yeah right nothing to beat here, agree. I think i achieved the main goal of winners from baseline a bit better than the chinese guy. Sure i had less winners in total numbers, but considering the set was short (in both games but especially points) and i made also way less errors we can call it a day. About average rally lenght, yeah i cant see how i would achieve less than 6 while sticking to all rules and detracting aces/df. Maybe going for 100 power in serve and groundstrokes and 0 top spin could lead me to area between 5-5.5, yeah thats possible.

And on top of that, the CPU did way too many net rushes which denatures the result. He didn't do that out of "desperation", but because you camp on the wall and thus do short defensive strikes quite often.


I just rewatched match until 4/0. I dont think i "camped" on the wall. As you can notice, during the rally only ankles of my player are not visible anymore, and only occasionally i go as far as knees aint visible (this happened maybe in 3-4 up until that point). NOw considering i could go as deep/backward as my entire player is not visible at all tells us i was being at max half way to wall (area in the middle between wall and baseline). 90% of the time this was happening only on return games and only if i was dealing with 1st serve. Its just natural i wont step into the court when im dealing with 210+ serves on the line and hitting average return at best, would not make sense to go into the court, especially knowing i can expect 143kh forehand bomb right next shot. Thats how works online. Im still one of the very few people you would find online, both ITST and MG tour who stay actually THAT close to baseline despite unfavoruable conditions. ITs related to drops too basically, since i were hugging the wall, i could not be able to catch all of CPU drops he tried. Same with short accels and normal short shots. If i were very deep in the court i would/should suffer from double bounces or errors from my racket, but that was not the case at all, i was always there to reach them rather comfortably (hitting very good passings from them), implying i was far from camping at the wall, especially considering cpu had very low spin stat so shots should bounce very high in such cases. About short strikes, forcing CPU come to net. From what i rewatched, most (roughly 70%) of those shots CPU feet were right on the baseline or just slightly before it. Not sure thats exactly opporutnity CPU should look for when deciding if to go to net, especially if we are talking about BASELINE playstyle. The nature of the game mechanics also makes those weird misshits which lands short possible. By playing in really high intensity you reach out more balls, giving yourself bigger chance to play such shots even though they should maybe end up in the net, nothing i can prevent from happening that. Also if i would play even closer to baseline during obvious offensive stance of CPU( and in due time i would be able to move very closely once i got used on such settings with enough playtime), it would again only create even more of those weird short shots due to playing high difficulty balls (created by high pace shots from cpu and sligth deviations from sweetspot as its nigh impossible to hit constant perfect sweetspot from such speedy shots and high speed stat). SO we cant really have playing extremely close to baseline and not hitting weird shots. THats just how the game works. I even tried elite controls if they are more punishing about this but they were not. If anything elite controls was generating even bigger landing zone, creating more floaty or short balls in the process.




- You have insane reflexes :P (I measured them as low as 117ms : I measured Pros on a real court at around 160-200ms)


Thanks. Yeah i think its caused by the fact that from the day i purchased TE2011 i played only and only with previews off, including offline before i discovered online part. I always thought it to be way too invasive and immersion breaking. By playing without i must have developed good hand eye coordination and general focusness for the ball.

- Slow AutoPos + Simu Controls, in comparison to Medium + Elite, allow both easier defense & attack, mostly because they are less exigent on the up/down positioning (this was expectable)


i still thought in Medium auto could not offset the Elite more punishing deviation from sweetspot and ESPECIALLY trying to hit ball on the rise, getting to big yellow landing zone



- You position yourself well but not very well with Elite controls : strangely, you walk back easily, but often you don't step in as you should, given away some precious ms to your opponent (eg: you likely missed a couple of direct winners with the short acceleration)
- You often prepare your strike late for the Medium AutoPos and thus don't enjoy all its benefits (ie: you start to press the button like if you were playing with Slow AutoPos, which I was kinda suspecting)


Yeah agree, getting my sweet spot right takes time on this setup. The fact i was very often "overrunning" shots or could not get right small steps is due to the very high speed and run acceleration. Simply aint used for such huge speed and reactivity since i play with much lower values in these two stats. Tho with time i would adjust

Thus said, I think it shows the differences between the 2 setups are a bit smaller than what I thought, but they are here. :)


Agreed there is difference. Slow autopos + Sim still gives me much better sensations especially from offensive playstyle


Now to testing another setup. Few observations.

This was my first try and most likely the last and I think i wont undertake another set with this specific steup, as it was very tedious, im sure you will understand. As i thought before trying, it was much more gruelling, tedious and overally unrewarding trying to attack from baseline. I stayed with IC5 level, since going higher would be senseless, result would be same but with gradually higher average rally lenght. BIggest issue was, as expected, a unpleasant snowballing effect of stamina due to several very long rallies. It was most apparent when i was leading 5/0, trying to stay true to offensive and risky playstyle.It felt frustrating that one tries to attack left and right and all it needs for opponent is to hit one long spinny shot to restart the rally, while you deplete your entire short term stamina. Even more frustrating is however, that you are punished for that even next rally/rallies, because your stamina does not recover to its fullest, further hindering your offensive attemps in next rallies, snowballing. Maybe a fix for this would be to replenish stamina to full after every rally, no matter how long or short (while still to max of long term stamina) and maybe boost a bit recovering long term stamina every odd game (during changeover).
Attacking from the baseline were often just pointless since combinations of clay, defender style and higher top spin puts too much difficulty penalty on my accels/short accels. Shoulder height accel bonus was also neutralized by the fact that to hit it i first need to step back, thus taking away time to charge longer, leading to not very accurate shot, while at the same time giving more time for cpu to reposition himself due to longer ball travel time. What was usually effective was wide serve and finishing the point into open court. This was however conditioned by cpu going for short and wek down the line return. And of course net play was effective but i did not want to overuse it. Also trying to lure cpu into the court with deliberate short shots and then finishing with accel seemed efficient to an extent. Despite doing fair share of baseline winners.
I honestly cant think how to get to lower 8 for average rally lenght. I had even 90 power and 0 top spin to boost my pace.

https://imgur.com/a/uG0BU6g

2020-05-12.11-58 - pidzi vs G.Lad 6-3.dmo
(107.39 KiB) Downloaded 27 times


About testing TE4. I will gladly do so once the occasion arises. For now i dont own TE4 copy, for reason you mentioned (that gameplay is not stabilized yet, basically its still tailored more or less for cpu vs cpu matches), but i wanted to get a copy once its gameplay is close to final version. However if there was a chance to give you some suggestions or feedback based on testing during lets say early access i might get the copy sooner. We will see how it turns out.

All being said, these are all very intersting points of view and observations and i have to say im glad we are having these tests, while both giving their own points of view and observations and opinions about things.

Have a nice day!
pidzi
crazy of the little yellow ball
crazy of the little yellow ball
 
Messages: 68
Gaming Since: 21 May 2011, 21:32
Location: Slovakia

Re: It is impossible to hit a winner [with the ITST Mod]

Postby manutoo » 13 May 2020, 06:15

rharm2,
the deep counter is a bit less effective when you play with the Elite controls, because it'll land more often out or short, but only if you don't have the Precision noticeably above the Power.
I unfortunately cannot change that without rethinking some stuff and risking altering the balance on other cases.
However, it's already fixed at 90% in TE4 because of the more realistic physics ; ie: 90% of the time the last instant shots are slower & shorter or just out or in the net.

pidzi,
1) you're right, I over-exaggerated with "camped on the wall". You were just 50-100cm too far from the baseline and like you said, off-center strikes land shorter ; the main issue here is the CPU being stupid with his net rush decisions (I already tuned that in TE4 and I'll tune it more depending on reports :) ), and thus we cannot use him to test the attack efficiency in the game. He's only good to test the defense, and there's already some limitations (eg: no anticipation, doesn't adapt his distance from baseline, etc.).

2) For your last test, rally length was fine, but I'd have preferred if you could have keep the net approaches significantly lower (ie: under 10), although some rallies call for a net approach.
I'm short on time for the next couple of days, so I guess I'll be able to watch it only at the end of the week. I'll come back to you, when I do. :yes:

3) For TE4, once I will have ironed the most annoying gameplay stuff (ie: the points that keep being reported by the users), I'll publish a demo so anybody could try the game. :) (that may come within a couple of months, if sh*t stops a bit to stack up on my door :fear: )
ManuTOO
== Mana Games ==

>> I don't answer Private Message, except if it's really a _Private_ topic <<
User avatar
manutoo
Game Author
 
Messages: 18689
Gaming Since: 24 Jan 2004, 15:38
Location: France

Re: It is impossible to hit a winner [with the ITST Mod]

Postby rharm2 » 13 May 2020, 11:04

Thanks for the swift reply manu, I am glad we will see that edited in TE4! :D

I think it is best to please ignore pro10, people like him make me understand why you bear a grudge to ITST, there is only a select few like that here. So I'd suggest you ignore him in the future as he is talking complete rubbish :!:


Have a nice day

rharm
rharm2
crazy of the little yellow ball
crazy of the little yellow ball
 
Messages: 196
Gaming Since: 25 Feb 2020, 00:32
Location: United Kingdom

Re: It is impossible to hit a winner [with the ITST Mod]

Postby manutoo » 15 May 2020, 09:26

Pro10,
if you want I answer your messages instead of deleting them, you should first read what I wrote in my previous messages and stop posting non-sense that directly contradicts what I have done & said. :sad3:
ManuTOO
== Mana Games ==

>> I don't answer Private Message, except if it's really a _Private_ topic <<
User avatar
manutoo
Game Author
 
Messages: 18689
Gaming Since: 24 Jan 2004, 15:38
Location: France

Re: It is impossible to hit a winner [with the ITST Mod]

Postby manutoo » 21 May 2020, 19:51

pidzi,
sorry for the delay to get back to you, I always have dozen of stuff piling up at the same time... :sweat:

So I tried your .dmo, but strangely it desyncs a lot, for at least 1 whole game each time, so I could see only 4 games (after 2/0 till 5/0 and then 1 more).
My 1st thought is that you have a Modded anim lying in your Data folder, but as it works for some games, I'm really not sure it could produce such effect. Anyway, if you ever do more .dmo, please check for any modded stuff just to be sure it doesn't come from that. (I got a bunch of .dmo for TEM2 from Bug Reports and they worked well, so logically, TE2013 .dmo system should still be ok)

I'm going to give you some thoughts that you could maybe use to tune the ITST Characters (or at least establish some bases for future testing with TE4).

So from the little half I have seen, it looks like you couldn't do winners from the baseline and did a lot of net rushes (like the stats showed).

However, I noticed you didn't play with 100% Power / Consistency / Precision, but with 90/92/99. By doing so, you likely not impair your defensive skills or just a tiny little bit (from the loss of your strike depth), but even the loss of power can give you more time to reposition yourself so it can be good for defense.
On the other hand, the +9% on Precision doesn't help much on the attack, but it helps a lot on defense by allowing to do last instant strikes whose Aiming Zone still stay fully in the court (especially with the Simulation Controls).

I wonder if you have impaired consciously or not your attacking strength, but it made harder the goal to make winners from the baseline. (the +9% would have been very good for the Short Acceleration before the last update, though)

One important thing : if you use the Elite Controls with +9%, your precision will be already above using Simulation Controls with +0% (and +8% is a tiny bit under). I noticed you often use Precision above the Power with the ITST, although I don't have a complete image of what has been done there. But when I promote the Elite Controls, I always have in mind a +0% Precision. ie: last instant shot = big aiming zone = chance to be fault = attack was a winner ; or = chance to be short = next opponent strike can be a winner if he steps in correctly.

So you often misposition yourself (ie: hit with the body) and from your controls (Simu + Precision skill) you're not much punished for it. Personally, wanting to make a better tennis simulation, that annoys me a bit. (TE4 is already much more punishing for that)

I checked my reflexes with this site : https://humanbenchmark.com/tests/reactiontime . I'm still on average at 206ms, and my best is 184ms. As far as I recall, this is around what I was when I did the Incredible challenge in 2014 (I thought my reflexes where decaying, but it's only when I'm really tired :) ).

Anyway, could you check your reflexes with this site ? If you get around 120ms as a best result (or better) (ie: it means this site reflects correctly your reflexes in-game), could you ask the top ITST players to test their reflexes and report them here ? Could you try to see if there's a correlation between their reflexes and their defensive abilities ? (Note: they should not cheat to get the best results possible ; I need real data to get a real idea of what's going on and how to fix it :blackeye: ).

One crazy thought : I guess your reflexes are faster than mine by about 70ms, while the IC-5 reflexes are better than the IC-1 one by only 50ms (and there's no bonus for the Speed & Tonicity skills ; the MaxSpeed bonus kicks in after ~6 meter, and it gives a 20ms advantage after ~7.5m) ; however, I can do baseline winners against the Incredible-1 (with Elite controls) and it seems you barely can against the IC-5, so it may show you have trouble finding your best-attacking potential (at least because of your skill choices, and possibly because of the Controls mode), while you have obviously super maximized your defensive one. Although of course the consequences of reflexes on attack & defense might not be exactly equivalent.

So I'd really like to prove that (or prove I'm wrong :blackeye: ) by tuning the CPU AI to make winners against you (without cheating, obviously ; only using your level of reflexes). Unfortunately, I really don't have the time to do that for TE2013, but maybe I could for TE4 in a few months (one can always hope :aureola: ).
ManuTOO
== Mana Games ==

>> I don't answer Private Message, except if it's really a _Private_ topic <<
User avatar
manutoo
Game Author
 
Messages: 18689
Gaming Since: 24 Jan 2004, 15:38
Location: France

Re: It is impossible to hit a winner [with the ITST Mod]

Postby pidzi » 21 May 2020, 23:37

Yo,

good to hear from you again.

sorry for the delay to get back to you, I always have dozen of stuff piling up at the same time... :sweat:

NO problem here, i get it you have a lot to do.

So I tried your .dmo, but strangely it desyncs a lot, for at least 1 whole game each time, so I could see only 4 games (after 2/0 till 5/0 and then 1 more).
My 1st thought is that you have a Modded anim lying in your Data folder, but as it works for some games, I'm really not sure it could produce such effect. Anyway, if you ever do more .dmo, please check for any modded stuff just to be sure it doesn't come from that. (I got a bunch of .dmo for TEM2 from Bug Reports and they worked well, so logically, TE2013 .dmo system should still be ok)

No idea what could cause this. Its complete unmodded vanilla game i have installed on drive. The only thing which comes to my mind is it could somehow interfere with my ITST Te installed on the different drive, maybe it somehow intertwines in some hidden files, dont know.



However, I noticed you didn't play with 100% Power / Consistency / Precision, but with 90/92/99. By doing so, you likely not impair your defensive skills or just a tiny little bit (from the loss of your strike depth), but even the loss of power can give you more time to reposition yourself so it can be good for defense.
On the other hand, the +9% on Precision doesn't help much on the attack, but it helps a lot on defense by allowing to do last instant strikes whose Aiming Zone still stay fully in the court (especially with the Simulation Controls).

I did few quick tests, looking mainly at two things. First i did CPU vs CPU with both IC5 difficulty. First i set their power and precision on equal value for both wings and did 3 sets. Average rally lenght was 7,03 (6,5; 7,5; 7;2). Then i did another 3 sets but with both wings having +8 on for precision. Average rally lenght was 5.7 (6; 5,9; 5). Sure, the sample is not very big, but i would expect it would be at least clsoe or even +0 would have less average, reading what you said about it. Since the difference was pretty big tho, even if i did 50 sets for both im pretty much positive that +0 would fall short opposed to +8 precision. Im positive about it because i did thousands of simulating tests already while working on the mod, and higher precision always produced better results in both average rally lenght and overall sensations you get from attacking attempts. ANd now its even more pronounced as last TE update rendered short accels less effective in power and precision when precision is +0 and less. So i have to disagree that having more precision over power is worse than +0. Disagree about last moment shots making defense with higher precision better to the point that it prolonges rallies. This is when i looked at the second thing. With previews on i first did a few hits with +0 precision, and intentionally hit with "last moment" shot on the full run and did same with +8. The difference was pretty much negligible. BOth cases could never hit out with green landing zone, while yes with +0, you have a bit bigger landing zone than with +8 but still both reaching outside baseline. Difference is very little to matter unless its 50+ shot rally. You can do some extensive tests too to see how it fares.
In ITST settings we can easily spot the discrepancy in effectiveness of +0 and lets say +7 precision. We have characters designed in a way where one wing usually has higher precision than the other. SOme chars have lets say +4 on both wings. There is an apparent short acceleration inefficiency with +0 precision wings, no matter if its forehand or backhand. I mean look, we can still hit many winners from baseline with normal accels. The main gripe we have now is simply with the last update which definitely rendered short accel quite inefficient with +0. Before the update, we did beta test for 1.20 ITST mod, and it almost felt absolutely perfect, everyone was having great sensations about the game. Then after update, while everything stayed same, short accels got hit by nerf in some settings. We always used short accels as both finishing shots and set-up shots. Now finishing with them feels almost impossible even if inside the court and good positioning (as you could see yourself with my tests here that nor me neither cpu could hit winners from it). And thus i dont believe that +0 is the way. You also have to realize we are able to run down short accels with just average 66 speed and run accel to average 88 fh power and average 85 bh power in the roster. Thats more or less +20 in favor of speed. I could not emphasize more that before update it was pretty much perfect.
If its worth it anything. In our settings best players can still beat IC10 6-0 with fair share of winners. I will attach a demo of very recent attempt from one of our members. IF you have will you can watch it. I myself recently managed to beat it 6-3. That said i can imagine we could do this with vanilla settings. AT least not without exploits.
What i was actually thinking is that not precision, but if lower consistency could not make defense less capable. YOu said that consistency (among other things) governs also depth of shot. Maybe going down with consistency could render defensive shots shorter, and thus easier to attack from ? Whats your take on this idea ?
QUICK EDIT: I just did i few quick tests too. Same cpuv vs cpu settings as above but i reduced consistency to 65 on both wings. The result was really suprising for me. The average rally lenght went down to average or 4,5 (for +8 precision on wings) and to 5,7 (for +0). That is more than one on average rally lenght. Again sample was not big but still 6 sets clearly showed way downward trend with much lower consistency. So could it really be thta lower consistency (besides less effective short accels/shots overally) could help to actually lower average rally lenght due to more shorter balls from defense and thus more chances to actually step into the court and finish point quickly ? Whats your opinion about it mate ?



I wonder if you have impaired consciously or not your attacking strength, but it made harder the goal to make winners from the baseline. (the +9% would have been very good for the Short Acceleration before the last update, though)


I did set +8 precision intentionally, based on my experience from both thousands of online matches and even more of CPU vs CPU tests, having higher precision over power always yielded better results, always. But it was pretty much said above


So you often misposition yourself (ie: hit with the body) and from your controls (Simu + Precision skill) you're not much punished for it. Personally, wanting to make a better tennis simulation, that annoys me a bit. (TE4 is already much more punishing for that)

I also did (again just to be sure) quick test about bad mishits with elite controls. I was intentionally hitting with bad positioning (around 0,30-0,60 sweetspot value in dev mode) top spin shots or normal shots, and these shots were still very safe.



Anyway, could you check your reflexes with this site ? If you get around 120ms as a best result (or better) (ie: it means this site reflects correctly your reflexes in-game), could you ask the top ITST players to test their reflexes and report them here ? Could you try to see if there's a correlation between their reflexes and their defensive abilities ? (Note: they should not cheat to get the best results possible ; I need real data to get a real idea of what's going on and how to fix it :blackeye: ).

I did 4 tests in total. Best i got was 126ms, worst was 221. It mostly varied in 170-200 zone. All four together averaged at 179ms. Im pretty sure there are some guys with higher reflexes. I can ask some guys to do the test and edit later.
QUICK EDIT: https://humanbenchmark.com/users/5ec6fe ... actiontime , Thats reaction time from Isniper, one of the TOP3 best playing players right now. I can ask another 2 or 3 guys to take this test. I think they would be around this reaction time too.

One crazy thought : I guess your reflexes are faster than mine by about 70ms, while the IC-5 reflexes are better than the IC-1 one by only 50ms (and there's no bonus for the Speed & Tonicity skills ; the MaxSpeed bonus kicks in after ~6 meter, and it gives a 20ms advantage after ~7.5m) ; however, I can do baseline winners against the Incredible-1 (with Elite controls) and it seems you barely can against the IC-5, so it may show you have trouble finding your best-attacking potential (at least because of your skill choices, and possibly because of the Controls mode), while you have obviously super maximized your defensive one. Although of course the consequences of reflexes on attack & defense might not be exactly equivalent.


I would not focus too much on the fact i could not hit many winners as a whole. As you could see i could still winners from baseline with normal accels when i was positioned well. I would focus on the short accels that did not provide me almost no real advantage even when hit perfectly, right on the line. This is all im trying to prove here, and measly difference between 92 and 100 power would not make it different. if anything i would do more errors because of bigger aiming zone.

Im looking forward for your opinions.
Attachments
LastGame.Dmo
(66.05 KiB) Downloaded 18 times
pidzi
crazy of the little yellow ball
crazy of the little yellow ball
 
Messages: 68
Gaming Since: 21 May 2011, 21:32
Location: Slovakia

Re: It is impossible to hit a winner [with the ITST Mod]

Postby manutoo » 22 May 2020, 12:01

pidzi,
1) If your vanilla TE is in another folder and you didn't specify another Mods folder within its Mods.ini, it cannot interfere with your ITST Modding on another drive or folder. So it might be a bug... :fear:
Or maybe you have tweaked your Tennis.ini ? (although it shouldn't influence the .dmo replaying, as normally all meaningful game parameters are saved in the .dmo )

2) I had checked the Precision formula too quickly and didn't notice there was a 2nd line in it. :sweat:
So it's actually : Elite +19% = Simu +0%.
Also I wasn't clear in my previous message, but I meant in Free Mode (ie: when the skills aren't remapped, like in the ITST Mod and normal offline mode).
If you did your tests with the Fair Mode still on, then you actually got +5.4% ( = +9% * 0.6 ).
So a true +9% is noticeable, but less of a game-changer than switching from Simu to Elite.
I took a few screenshots in a standardized test to clearly view the differences, using the normal strike with a preparation time of 0.1 seconds (ie: if you play online with a Delay above 100ms, this strike will be seen as a no-button strike), 100% Consistency, and the ball not exactly in the center of the cords so the Sweetspot is 0.2 in Simu which gives 0.25 in Elite :
TE2013_Simu+0.jpg
Simu+0

TE2013_Simu+9.jpg
Simu+9

TE2013_Elite+0.jpg
Elite+0

TE2013_Elite+9.jpg
Elite+9

TE2013_Elite+17.jpg
Elite+17


I think it clearly shows the Simu+9 has a significantly lower chance for the ball to go out, and it can't be as short as the other setups.

Oddities :
- perfectly centering the ball will make it deeper and thus gives more chance to make it out
- a lower Consistency makes the ball shorter and thus removes the possibility to send it out ; this means once you're under 80% Consistency, the strike has 0 chance to end out with Simu+9 while still a decent chance to be out with Elite+0 (ie: it's a game-changer in term of defense, but has little effect on the attack)

3) They are too many things going on in determining the Rally Length so it's hard to conclude something only with it, especially in CPU vs CPU matches

4) I'm not sure where you are in terms of Consistency for your ITST Characters, but as outlined above, lowering it may benefit defense more than the attack, especially if the user isn't comfortable with stepping on or in the baseline.
Also, you get a +1km/h Speed bonus for the Short Acceleration per 20% of Consistency.
Your CPU vs CPU test is indeed surprising ; one thing that comes to mind is that the CPU doesn't position himself perfectly up/down (ie: sometimes I make him step inside the Danger Zone to simulate a little mistake) and thus could be making more mistakes with a lower Consistency. You can check the error stats to confirm or infirm that.
Else, it's quite possible that the shorter balls gave more attack occasions or allowing to rush the net more. Again, the winner stats & net rush might help to determine this. Also here as well you should evaluate if the common ITST user can exploit correctly the shorter balls without the help of the Danger Zone (or at least the Impact preview).

5) I quickly checked my reflexes in-game and I can get as low as 100ms (without cheating :P ) but also as bad as 267ms... :shock:
So I guess the main advantage you have over me is the stability of your reflexes, even if the online test shows only a ~25ms difference.
I'd still be interested to get an idea of the reflex levels of your fellow top ITST players, even if the correlation between the game & the online test isn't that great.
Or maybe they can test in-game : need to go in warm-up, send back the 1st ball to the opponent, not move, wait for the opponent to strike left or right, start to run left or right correctly, and then check the replay in the step by step mode to count how many frames it took you to react after the 1st frame where the ball changed of direction and then you can multiply it by 16.7ms. Note : in Step by Step mode (ie: replay is paused), B1 = advance 1 frame ; B2 = rewind 1 frame.

6) If the v1.2 felt perfect (or near of it) before the last TE update, it might be possible to tweak a bit the Character Skills to reach the same state.
So 1st :
ANd now its even more pronounced as last TE update rendered short accels less effective in power and precision when precision is +0 and less

You may have misunderstood what the update did.
1st, the strike precision didn't change except when the Precision skill was above the Power by 20% or more (I guess it's not the case in ITST).
After, the maximum speed bonus from Precision vs Power of the Short Acceleration was for +20% : 7.5km/h ; now it's 4km/h.
So if you have a Character with +10%, he had +3.5km/h which is now +2km/h.
There's also a Precision speed bonus which is +1.5km/h per 10% of Precision (ie: 0 at 0% and 15km/h at 100%)
So you can raise the Precision by (3.5 - 2.0) / 1.5 * 10 = 10% to get the same speed.
Also, the speed from Power is +3km/h per 10% Power (but only for low spin balls), so you can raise the Power by (3.5 - 2.0) / 3 * 10 = 5% to get the same speed. (with a heavy spinner, you'd have to raise it a bit more)
Considering this, I guess I'd try a 3% raise on both Power & Precision ; this would give a bit more speed on the normal acceleration and an acceptable tiny loss on the Short Acceleration.
However, maybe your main issue is the loss of the speed bonus on off-centered balls. For this, there's no easy fix.

7) The short acceleration is best with a high topspin because in such a case it's less slowed down when the ball is hit low. Your last try was with 0% topspin so it might be normal that the short acceleration wasn't that effective, even on clay. A 35% Topspin skill may give a different result.
ManuTOO
== Mana Games ==

>> I don't answer Private Message, except if it's really a _Private_ topic <<
User avatar
manutoo
Game Author
 
Messages: 18689
Gaming Since: 24 Jan 2004, 15:38
Location: France

Re: It is impossible to hit a winner [with the ITST Mod]

Postby pidzi » 22 May 2020, 18:49

Hey,

thanks for quick response, these are some really nice insights and suggestions.

1) If your vanilla TE is in another folder and you didn't specify another Mods folder within its Mods.ini, it cannot interfere with your ITST Modding on another drive or folder. So it might be a bug... :fear:
Or maybe you have tweaked your Tennis.ini ? (although it shouldn't influence the .dmo replaying, as normally all meaningful game parameters are saved in the .dmo )


I did not do any tweaks in .ini, i just installed and went straight on testings so i guess its only some sort of bug.

2) I had checked the Precision formula too quickly and didn't notice there was a 2nd line in it. :sweat:
So it's actually : Elite +19% = Simu +0%.
Also I wasn't clear in my previous message, but I meant in Free Mode (ie: when the skills aren't remapped, like in the ITST Mod and normal offline mode).
If you did your tests with the Fair Mode still on, then you actually got +5.4% ( = +9% * 0.6 ).

I believe i did them with free mode but would not wager my liver on that :D .

I think it clearly shows the Simu+9 has a significantly lower chance for the ball to go out, and it can't be as short as the other setups.

Yeah there is difference still i would wish it would be even bigger tho considering its really just last moment hit, it should be imho bigger landing zone. My two cents tho.

Oddities :
- perfectly centering the ball will make it deeper and thus gives more chance to make it out

Yeah one would expect it could be the other way around in this area

- a lower Consistency makes the ball shorter and thus removes the possibility to send it out ; this means once you're under 80% Consistency, the strike has 0 chance to end out with Simu+9 while still a decent chance to be out with Elite+0 (ie: it's a game-changer in term of defense, but has little effect on the attack)

I will have to test this further to see what is impacted more with lower consistency. SO far it seems its the opposite but as i said, i did smaller sample and only with cpu matches. I will update on this.

4) I'm not sure where you are in terms of Consistency for your ITST Characters, but as outlined above, lowering it may benefit defense more than the attack, especially if the user isn't comfortable with stepping on or in the baseline

We have average 82 consistency when i do mean value of both wings, and it varies from as lowest as 78 to as high as 85. So maybe upping those under 78 to 80 would do some kind of trick.

Also, you get a +1km/h Speed bonus for the Short Acceleration per 20% of Consistency.

Is this an actual threshold thing ? I mean if i have 79 consistency i get for example 120kmh short acc, but the moment i up it to 80 its 121kmh (not taking into consideration any other variable ofc) ? This would also mean that someone with 80 consistency gets the same speed as the one with 99 with all other stats completely same ?

Your CPU vs CPU test is indeed surprising ; one thing that comes to mind is that the CPU doesn't position himself perfectly up/down (ie: sometimes I make him step inside the Danger Zone to simulate a little mistake) and thus could be making more mistakes with a lower Consistency. You can check the error stats to confirm or infirm that.

Yes there were more errors in total in matches with lower consistency, winner stayed more or less on same value.

Else, it's quite possible that the shorter balls gave more attack occasions or allowing to rush the net more

Thats what i was thinking lower consistency could allows us to actually step into the court more for easier finishers.

5) I quickly checked my reflexes in-game and I can get as low as 100ms (without cheating :P ) but also as bad as 267ms... :shock:
So I guess the main advantage you have over me is the stability of your reflexes, even if the online test shows only a ~25ms difference.
I'd still be interested to get an idea of the reflex levels of your fellow top ITST players, even if the correlation between the game & the online test isn't that great.

I already posted mine which was those 179 mean value after 5 tests, Isniper got 147ms. ANother top guys, Richie was averaging around 190, same goes for Pablo, Mohd did the test and made 176ms. Hope this helps.

Or maybe they can test in-game : need to go in warm-up, send back the 1st ball to the opponent, not move, wait for the opponent to strike left or right, start to run left or right correctly, and then check the replay in the step by step mode to count how many frames it took you to react after the 1st frame where the ball changed of direction and then you can multiply it by 16.7ms. Note : in Step by Step mode (ie: replay is paused), B1 = advance 1 frame ; B2 = rewind 1 frame.

I will checkout this way and maybe ask others to undertake it.

6) If the v1.2 felt perfect (or near of it) before the last TE update, it might be possible to tweak a bit the Character Skills to reach the same state.
So 1st :
ANd now its even more pronounced as last TE update rendered short accels less effective in power and precision when precision is +0 and less
You may have misunderstood what the update did.
1st, the strike precision didn't change except when the Precision skill was above the Power by 20% or more (I guess it's not the case in ITST).


This time i would wager my liver :D, that short accels went down in power even for +0 precision. Since TE update came out just a few hours after we released 1.20, we still had in mind very fresh sensations from betatesting 1.20 and we definitely noticed that pace of the short accels was hit in downward trend even for +0. I would like to nicely ask you if you could maybe double check previous version and current one if all is ok. IF yes then it will be probably placebo, but could swear im positive about my impressions. Thank you in advance.

After, the maximum speed bonus from Precision vs Power of the Short Acceleration was for +20% : 7.5km/h ; now it's 4km/h.
So if you have a Character with +10%, he had +3.5km/h which is now +2km/h.
There's also a Precision speed bonus which is +1.5km/h per 10% of Precision (ie: 0 at 0% and 15km/h at 100%)
So you can raise the Precision by (3.5 - 2.0) / 1.5 * 10 = 10% to get the same speed.
Also, the speed from Power is +3km/h per 10% Power (but only for low spin balls), so you can raise the Power by (3.5 - 2.0) / 3 * 10 = 5% to get the same speed. (with a heavy spinner, you'd have to raise it a bit more)
Considering this, I guess I'd try a 3% raise on both Power & Precision ; this would give a bit more speed on the normal acceleration and an acceptable tiny loss on the Short Acceleration.
However, maybe your main issue is the loss of the speed bonus on off-centered balls. For this, there's no easy fix.

Very nice you provided some hard fact numbers here. THis could help tremendously in shaping the mod in right direction.

7) The short acceleration is best with a high topspin because in such a case it's less slowed down when the ball is hit low. Your last try was with 0% topspin so it might be normal that the short acceleration wasn't that effective, even on clay. A 35% Topspin skill may give a different result.


IN my initial tests i believe i played with around 30-35 top spin too and the difference was negligible compared to 0% spin. Tho i agree going to around 40-50, top spin starts to create some good angles on short accels, but then again it hampers normal accelerations with lower pace. SO its a tradeoff again.
pidzi
crazy of the little yellow ball
crazy of the little yellow ball
 
Messages: 68
Gaming Since: 21 May 2011, 21:32
Location: Slovakia

Re: It is impossible to hit a winner [with the ITST Mod]

Postby manutoo » 23 May 2020, 08:20

1) Last instant strikes in TE4 get a bigger zone, and more importantly can end in the net ( => likely a game-changer in term of defense ;) )

In a similar way, if the ball is shorter due to lower consistency, it means it'll have more chances to end in the net, so it'll make more sense in TE4.

2) It's not a threshold, it's linear (it's very rare that I use threshold), so the formula is : speed bonus in km/h = Consistency / 20

3) Considering your CPU vs CPU stat changes, I guess it's worth to make a try in a couple of online matches, to see the result (or at least Human vs CPU :) )

4) Ok, I guess there's no specific pattern associated with the reflex level

5) I just ran the debugger for you : the speed bonus is really 0 when the Precision is equal to the Power. BTW, if the Precision is under the Power, it's a handicap, and it means your Character with Power above the Precision (I guess you have some ?) got their Short Acceleration slightly faster, especially when not centering the ball, as the handicap is also removed in that case.

6) Since TE2009 the balance between attack & defense has always been a struggle, and thus I boosted the effectiveness of the short acceleration for players with high Topspin skill to compensate for the loss of attack efficiency on the normal acceleration. It's also relatively realistic as it's much easier to hit strong short strikes with high spin.
ManuTOO
== Mana Games ==

>> I don't answer Private Message, except if it's really a _Private_ topic <<
User avatar
manutoo
Game Author
 
Messages: 18689
Gaming Since: 24 Jan 2004, 15:38
Location: France

Previous

Return to Online Tennis Games

Who is online

Users browsing this forum: No registered users and 3 guests