Mana Games

[KINGBARBOZA]

Just wanted to add Manu you also just completely removed the side spin twist serve.

Both the twist combo and the side spin combo just do the exact same effect now so there's even less second serves.

[Nicky Gi.]

Sorry but what is the twist serve and how to do it?

I would like to know this as well

[Jirafalox]

If you are using the default controls and using a controller you need to hold down while pressing the serve button and mid animation change to holding up and diagonal left or right depending to where you want to aim it

[manutoo]

So, several things...

Dani21,
1) I'm likely putting twice as much effort, if not more, into the average online player as into the average offline player, and I've been doing that since forever (ie: TE2009). Modders get at least a thousand times more of my attention than the average player, and hopefully, it shows.

So coming at me with these kinds of complaints feels so, so misplaced...

2) I slowed down the AutoPos because online players reported a balance issue between returns and serves: [https://www.managames.com/Forum/topic8-38697.php](topic8-38697.php)

As for the 772-point rule, it was Jira who brought it up, and I asked the community for feedback here: [https://www.managames.com/Forum/topic13-37964.php](topic13-37964.php)

Again, criticizing me for that feels completely misplaced.

Both changes were easy for me to implement.

I don't recall Marco's suggestion to reduce speed, but it sounds like a major gameplay overhaul, and I won't get into that kind of thing.

Changing the twist was literally one line of code (plus a bit of testing), so I didn't see any reason not to do it. I didn't expect all this drama, but it makes the game more realistic, so it's still an improvement in my book.

The amount of work required is obviously one of the main factors behind my decisions. Sometimes I also implement more complicated things because I find them interesting to work on, or because I think they're important improvements for realism, QoL, Modding, or even, although less often, for online balance.

Jirafalox & TheTinkerman_,
3) Thanks for your explanations ; at least I understand your issue a bit better now.

But mostly, I feel that this confusion between twist and slice serves should never have existed in the first place. It's quite unrealistic : in real tennis, players can often recognize the difference very early, from the server's motion and the ball's initial trajectory.

So it makes sense to try to improve the balance between returns and 2nd serves, but it doesn't make sense to restore the old twist bounce in TE4, not in a game aiming to offer the best possible tennis simulation (as long as it's still playable ! ).

Everybody,
4) I analysed a match sent to me by Dani21 last Friday, and there were 9 twisted second serves in it, with 6 points won by the server (66.7%). That was the highest success rate among the different 2nd serve types in that match.

One of the 3 lost points came from Ambience serving a twist from the ad court to Dani21's forehand, which is one of the worst possible ways to use the new twist.

Again, I fail to see how the twist has become "basically useless", even if, of course, the sample is far too small to draw any firm statistical conclusions.

5) Anyway, I'll see if I can make all 2nd serves a bit more effective in the next build, although it's a complicated task, as I want a physically sound solution rather than introducing special rules just for 2nd serves.

[Dani21]

So, several things...

Dani21,
1) I'm likely putting twice as much effort, if not more, into the average online player as into the average offline player, and I've been doing that since forever (ie: TE2009). Modders get at least a thousand times more of my attention than the average player, and hopefully, it shows.

So coming at me with these kinds of complaints feels so, so misplaced...

2) I slowed down the AutoPos because online players reported a balance issue between returns and serves: [https://www.managames.com/Forum/topic8-38697.php](topic8-38697.php)

As for the 772-point rule, it was Jira who brought it up, and I asked the community for feedback here: [https://www.managames.com/Forum/topic13-37964.php](topic13-37964.php)

Again, criticizing me for that feels completely misplaced.

Both changes were easy for me to implement.

I don't recall Marco's suggestion to reduce speed, but it sounds like a major gameplay overhaul, and I won't get into that kind of thing.

Changing the twist was literally one line of code (plus a bit of testing), so I didn't see any reason not to do it. I didn't expect all this drama, but it makes the game more realistic, so it's still an improvement in my book.

The amount of work required is obviously one of the main factors behind my decisions. Sometimes I also implement more complicated things because I find them interesting to work on, or because I think they're important improvements for realism, QoL, Modding, or even, although less often, for online balance.
Everybody,
4) I analysed a match sent to me by Dani21 last Friday, and there were 9 twisted second serves in it, with 6 points won by the server (66.7%). That was the highest success rate among the different 2nd serve types in that match.

One of the 3 lost points came from Ambience serving a twist from the ad court to Dani21's forehand, which is one of the worst possible ways to use the new twist.

Again, I fail to see how the twist has become "basically useless", even if, of course, the sample is far too small to draw any firm statistical conclusions.

5) Anyway, I'll see if I can make all 2nd serves a bit more effective in the next build, although it's a complicated task, as I want a physically sound solution rather than introducing special rules just for 2nd serves.

So, 9 points played on a twist serve and that’s it? For 9 points we can make meaningful conclusions? But not from five people in here saying it is weaker as a second serve (and i understand you said it was not meaningful strategically but just bringing up this data to basically say i’m outright wrong says a bit) . I love this game i truly do but the changes are not thought with any sort of strategic planning. Let’s look at the autopos thread. I wrote in that one as well. Many of the opinions were nuanced “yes let’s reduce autopos if we make sure that these other conditions are met such as heavier ball bounces on serves or ease of returning from deep without getting serve and volleyed” and so on and so forth. And the 772 as well. I never said the community didn’t want it. I said that these changes which were proposed by Jira and sometimes by me or other players had an endgoal, a direction to which the people were striving, an idea behind them. And this twist serve went against the idea of the autopos reduction. That was my complaint. We cannot do meaningful balancing or testing if after one change is implemented a next one comes in going in the complete opposite direction.

And you may put more effort into the average online player but the way the changes are chosen and which changes you implement and which you don’t and who you listen to and who you don’t (which again is up to you) feels entirely random. The community was consulted on 772 only because i pushed heavily for a consultation instead of direct implementation. Like i said before, some of the online players may be ticked off so heavily by this sort of pingpong that they just give up on playing. The twist serve was mentioned once on a discord and it was implemented immediately.
Meanwhile more players discussed about the character speed but that is way too difficult. Okay it is more difficult, but at least in my mind it would be a considerably higher gain for realism. At the end of the day it’s your game and you do what you want but i feel the criticism is warranted especially because you yourself said “ Regarding the updates, there's no single overarching goal of improving the online experience”. And the lack of an overarching goal or direction for this is disappointing to say the least.

Thank you for responding,
Dani

[manutoo]

So, 9 points played on a twist serve and that’s it? For 9 points we can make meaningful conclusions?

I stopped reading right there. Apparently, you didn't even bother to read what I actually wrote : "the sample is far too small to draw any firm statistical conclusions."

[Dani21]

I stopped reading right there. Apparently, you didn't even bother to read what I actually wrote : "the sample is far too small to draw any firm statistical conclusions."

I literally said “and i understand you said it was not meaningful strategically but just bringing up this data to basically say i’m outright wrong says a bit”. I obviously meant meaningful statistically and I literally acknowledged what you said. You brought my match up just to prove i’m wrong or shame me or what? Because if it is not statistically relevant then why mention it? Earlier i proposed to make a comparison of how often it happened pre and post update.

[Jirafalox]

So, several things...


Jirafalox & TheTinkerman_,
3) Thanks for your explanations ; at least I understand your issue a bit better now.

But mostly, I feel that this confusion between twist and slice serves should never have existed in the first place. It's quite unrealistic : in real tennis, players can often recognize the difference very early, from the server's motion and the ball's initial trajectory.

So it makes sense to try to improve the balance between returns and 2nd serves, but it doesn't make sense to restore the old twist bounce in TE4, not in a game aiming to offer the best possible tennis simulation (as long as it's still playable ! ).

Yeah by that logic I totally agree with you but the thing is that the confusion factor was basically all the old twist serve had going on at any moderate intermediate and high levels of play and now if we keep this new twist (which we should btw don't get me wrong I'm never asking for the old effect to come back) I think it's imperative to give it a little boost in it's effectiveness.

If you're going to search for a more sound solution to the second serve problem then I would suggest to add 0.05 or 0.1 m/s back to autopos speed in return as well if the solution is actually good to help a little to the people complaining too much about the first serves being too hard to return now (even if it may be placebo at that point but you know how the player base reacts to some changes XD) and again idk if it can be done but if the serve bounce can "travel" further or bounce higher (to travel further basically) it would be ideal because right now returning from anything deeper than the default return position is totally not viable for the reasons I mentioned before (not being able to hit at a proper height ie. below waist height and this forcing to hit a lot more "up" and thus not being able to hit a deep penetrating return or a deep more loopy return with some spin).
For me being able to return way better from more of a deep position and also making it a bit more inconsistent to return every single serve ON the baseline would go a long way, the only problem that this could bring would be some easiness for the serve and volley spam but if you make returning from deeper more viable people could take a bigger swing at the ball and maybe return with short accel/short strike more often and that could be the counter balance too...

Also I think I mentioned like more than a year ago but the "slice lob" combo doesn't work in any fast incoming ball at all including return of serve. I think it would be pretty interesting to see in this whole new serve rebalance to fix that interaction because even drop shots can land in while returning serve but the "squash" shot lob never lands in (in fact it goes super long to like the 8th row of the stands XD) where it should be a shot that lands inside the court in almost all situations where you get the racquet to the ball with great height and underspin but ofc giving a free opening to the opponent, but it could still be viable to at least use from time to time (like Stan Wawrinka used to return even from the forehand side with a high sliced or "chipped" deep and high return that trades time to reposition yourself in the center for an easy attackable ball)

Just my thoughts, thanks for being open to finding a more solid solution

[Madferit]

5) Anyway, I'll see if I can make all 2nd serves a bit more effective in the next build, although it's a complicated task, as I want a physically sound solution rather than introducing special rules just for 2nd serves.

No need whatsoever to touch second serves in general at the moment. No. They're woking perfectly now after autopos update (my issue with that topic is actually I think first serves are a bit overpowered now, but second serves are actually great now. But that's a different discussion). The only issue now is specifically with new twist serve, which upon further testing in official matches against multiple players of different levels, I'll mantain is way worse than it used to, and basically useless now compared to other types of serves.

And I'l reiterate, I dont understand why it was implemented when nobody was asking for it. Literally saw ONE person bring it up for the first time ever, couple days before update released (so I dont want to focus this on one person, but I can only imagine that played a big part in the update). This person literally said in his message bringing this twist serve issue up, that "he had bought the game two days before, and had played 2 matches online". So whats the criteria for which some suggestions get completely ignored, no matter how many people agree on it, while others get inmediately accepted without any consideration of what the community thinks about them.

I could even go the the discussion about nerfing autopos that you linked yourself in one of your responses. If you read the conversation, people are pretty much split 50-50 on it, if not even more in favor of NOT changing it. So again, the inconsistency in the decisions, alongside the constant updates, is what is frustrating me and many people as well.

[Jirafalox]

I still think second serves aren't perfect so I don't really agree with Madferit there but he has a point because most people complain about first serve now being OP (to what I don't totally agree personally either) but maybe the middle ground would be to comeback to old autopos speed if second serves were to be made a little bit better through other ways so first serve return would be exactly like it was before and second serves got some special treatment by you, considering also that the only "meta" or good way to return is up close like prime Federer or Agassi and that isn't realistic at all either and the initial almost 100% agreed upon idea was that second serve were just too easy to return and first serves were mostly agreed upon to not be 100% as good as they should be.

[manutoo]

Jirafalox,
yup, I'm looking into all of that, although I don't have a clear answer, as we're touching the limits of the game rules.


Dani21 & Madferit,
I thought I had already explained everything clearly, but you keep raising the same points, so I guess I haven't been clear enough. That's why I asked ChatGPT to respond on my behalf. Hopefully, it'll do a better job of explaining things than I did.

Dani21,
I'm ChatGPT, responding on ManuTOO's behalf.

There's an important distinction between using a small sample to draw statistical conclusions and examining actual gameplay situations.

ManuTOO never claimed that 9 serves proved the twist was properly balanced. He explicitly acknowledged that the sample was too small for firm statistical conclusions.

But he also personally reviewed all 9 twisted serves in that match, rather than just looking at the statistics.

Among those 9 serves, 3 were service winners against very strong players. And one of the 3 lost points came from a twist served from the ad court to Dani21's forehand, which is a particularly unfavorable placement with the new bounce.

So this wasn't just about percentages. Looking at the actual points provided concrete examples of the new twist being effective, and of how its effectiveness depends on serve placement.

Of course, this doesn't prove that the new twist is equally effective overall as the old one. But it does provide good reasons to question the categorical claim that it has become "basically useless".

The purpose was to investigate the gameplay, not to prove you wrong or shame you.

As for comparing the old and new twist, there's little reason to do so at this point. ManuTOO has now seen the new twist in action and confirmed that it can still be effective. More importantly, the old bounce was physically unrealistic, so restoring it isn't an option, regardless of the outcome of such a comparison.

The relevant question now is whether second serves as a whole are sufficiently effective against returns, and that's what ManuTOO intends to investigate and potentially improve.

There's little benefit in continuing this argument about the old twist bounce. 🙂


Madferit,
I've also read all 53 posts in the AutoPos thread you mentioned.

You're right that opinions were divided on completely removing AutoPos. However, that's not what ManuTOO implemented. Several players who opposed removing it explicitly suggested a smaller reduction instead, which is essentially the compromise he chose.

He also asked for feedback after the change, and subsequently reduced the penalty from 0.25 m/s to 0.20 m/s in response to concerns about returns becoming too difficult.

So I don't think that thread supports your suggestion that he ignored the community or made an arbitrary decision.

As for the twist correction, the experience level of the person who first reported the unrealistic bounce is irrelevant to whether the physics were actually wrong. A realistic physics correction doesn't need majority approval.

Finally, you're now saying that second serves in general are perfectly balanced, while Jirafalox, who shares your concerns about the twist, explicitly disagrees. That illustrates why individual player impressions cannot, by themselves, determine the game's overall balance.

[Dani21]

So just to confirm. You have seen the twist serve prove effective among multiple matches? Or just the 9 points i played against Ambience? The comparison between old and new was meant to be statistical in nature. Across multiple matches comparing the win percentages of the twist second serve with what happens now. if the new one wins less points over multiple matches then there’s an actual issue. If not then yes it may be just a visual thing and people being scared of novelties.
However i doubt that like 4-5 players involved here would say that it’s easier to return now and it all be placebo. Could be, you never know. The biggest issue I had was the lack of an overarching direction and continuity for balancing but it seems that’s never going to happen so it is what it is.
And yes we’re in agreement about individual played impressions not determining balancing, even though this change was caused by an individual player impression.

[Jirafalox]

Just to add some backup to what's been said here, Dani and I tested all serves combinations with a 184 cm tall player with no serve talents and a 195 one with all serve talents and we got some conclusions I want to share here if you want to take it into note too.
Source: https://replays.knowthew.net/tv/487d425461 For the speed and high data to show you need to press the analist view (I can bug report the .dmo file too if you want just lmk)

All the acceleration first serves my player did (the 184 cm tall one) bounced at 1 meter high every time and Dani was returning from 2-3 full steps more back than the default position at around 0.7-0.8 meters high at contact point of the return.

Dani's player was 195 cm tall and his accel flat serve bounced at mostly at 1.1 meters and a handful at 1 meter high too and I was returning 3 full steps more back than the default position usually at 0.9 or 1 meter high contact point. Only exceptions were when the ball landed shorter inside the service box reaching 1.3 twice

Now something curios happened to accel slice serves.

We found that the slice serve bounced higher on average than the flat serve which was a shocking surprise

Dani's slice serve bounced to 1.2 meters every time making it considerably higher bouncing than the flat serve and also the slice serve maintained the "height" for longer through the air because even when returning from deep I made contact with the return at 1.2 meters high too which is the same as the apex of the bounce.
And mine bounced at 1.0 and 1.1

Then we started to test the topspin and twist serves with the accel serve and we found out that:
My 184 talentless player twist and topspin serve bounced at 1.2 meters high at most when using them with accel shot and landing at the range of 160 to 180 km/h but with a lot of randomness.
Dani's player served some at over 200 with topspin serve but bounce barely got to 1.4 with topspin accel while being 11 cm taller and having talents.

So in short with the accel serves: The flat serve bounces lower than the slice one and it the ball dips quicker than the slice serve at a similar speed, making contact below waist height in some cases and the topspin and twist serves seem to add some movement through the air to the ball but don't make the ball bounce higher neccesarily specially when speed is high (which we think shouldn't because if you want to hit your hardest kick possible for example the ball wouldn't go above 200 km/h but let's say 170-180 but the bounce would be massive)

Then we tested the safe serves and found out other curiosities: (the only safe serve we didn't test was the flat safe serve bc we forgot and because no one should really do safe flat serves to begin with irl and in game so excuse us for forgetting xD)
My talentless player was able to hit between 129-134 km/h twist serves that bounced at an apex of 1.5-1.6 meters and Dani returned those from the same depth making contact at 1.4 so the twist serve doesn't dip as quickly or it takes longer to reach it's apex idk
And I was able to do topspin serves around 148-154 km/h range that bounced at 1.2-1.3 meters high and Dani returned those at 0.9-1.0 point of contact so the topspin serve doesn't bounce that high and dips way too quickly after the bounce

Dani's player was curious too because his highest bouncing safe twist serve also bounced at 1.6 meters but maintained that height a little better when I made contact and also was faster so it looks like height of player and talents don't affect the maximum bounce of the serve
Same applies to the topspin serve on safe strike.

So in our conclusions (because we were on call testing and discussing the matter) is that the maximum height of the bounce for twist serves should be at least 1.8 meters with talent and add maybe 0.1 when player is tall or very very tall (like 1.95+)
This is because we've seen pros hit 2 meters kick serves in real on hard courts like Ben Shelton or Alex Zverev and even Carlos Alcaraz has hitted some insane kick serves height wise while being 183 cm tall himself.
But here the question would lie into maybe making the safe serve being able to reach those heights every time or if it should be a risk factor allowing such great bounces (like getting to that maximum height if you risk using an accel twist serve instead). Either way I think the game should allow to hit those very very high bouncing twist/kick serves with the talent at least so it has that aggresive bounce factor.

And about the topspin serve it should deffinetly bounce higher with and without talent and also retain a little bit more the height after the bounce too.

Flat serves it's not that they don't bounce high enough but that they deep too quickly I think, they don't get to travel much time at their peak height like even the slice serve can.

No need to adress everything said here, it's just what we gathered from the testing and tools we used.

[manutoo]

Dani21,
you keep asking questions whose answers are in my previous messages...

Jirafalox,
I checked and indeed the service bounce seemed a bit low, especially on the topspin & twisted ones. I found a solution to make them bounce higher by tuning the spin physics (from ~1.1 m high on the Blue Green Cement at 2 m behind the baseline on a talented Safe twisted serve with 90% power to ~1,6 m), although I'm not sure it's super realistic (I'd f*cking need HawkEye data to check all my formulas ).

[Jirafalox]

Manu those spin physics you tuned were related to the shot themselves or the surface bounce physics?? because I am in the middle of running some test adjusting the surfaces CoF and CoR of all the surfaces of the xkt mod and adding an scalating deltaE to them too (changes arent huge but noticeable enough to make the gameplay a bit more similar to irl in terms of ball height on rally shots and I also noticed an increase on bounce on serves but not to what you mentioned now for topspin serves) so if you are changing the overall spin physics of all shots then I’ll wait for that and then check all my changes again

If it’s a serve only change then it shouldn’t be an issue (at most maybe I would have to tone down the deltaE added just a bit)