Mana Games

[Ambience]

Hi Manutoo and community,

First of all, a big thank you Manu. TE4 is by far the best tennis game out there, and the care you still put into it is rare, especially as a solo dev. I know you're mostly done with big features and are now on tweaks and bug fixes, which is totally fair after so many years. I really appreciate that you still read everything, answer questions and claims, and take the time to explain your decisions. I know it isn't easy to deal with all of us, and we can be pretty annoying sometimes.

I fully understand that most players are offline or World Tour, and that you build the game for them first. That makes total sense. What I'd like to add is that online and high-level players can be useful early testers for everyone. We play thousands of matches against humans who exploit every small detail, so we often notice a balance issue before it shows up offline, where it eventually affects the AI and the overall feel too. The goal isn't to make the game "online first". It's to help keep it as realistic and balanced as possible in both modes.

That's why I'd like to suggest an optional test step before some gameplay changes go live:

1. A small group of volunteer testers (offline and online players, modders, mixed levels) could try a build a few days before release. No extra work for you: just send it to a few people, and we come back with clear bug reports, match logs and short, specific feedback on one mechanic at a time.

2. Polls as input, not orders. Like we did with the 772 points topic and the next-year Tour suggestions, big gameplay tweaks could get a quick community poll before they go live. The final call obviously stays yours, but this way a change isn't shaped by the few players who post the most or the loudest. Everyone gets the same voice, offline or online, casual or high-level, and you get a real picture of what the whole community thinks instead of a handful of opinions.

We've already done something like this: for the 772 points change, players ran a test exhibition themselves before it went in, and it worked well. I think a small, organised version of that could save you some back-and-forth after patches.

The twist serve change is a good example. I know you tested it and the bounce itself is more realistic. From what several of us have seen, the issue seems to come from the interaction with AutoPos: the new bounce now moves the ball toward the returner, so it ends up easier to return in both modes. Your own idea of adding a bit more aiming imprecision from the spin sounds like it could be the right direction, and that's exactly the kind of tweak a small test group could try for you and report back on quickly, with bug reports included.

So I'd like to invite everyone, offline and online players and modders alike, to give their two cents here, ideally short and specific. And Manutoo, if a small volunteer test group sounds useful to you, in whatever form works best for you, we'd be happy to help however we can.

Thanks again for everything. Let's keep making this the best tennis game ever

[Jirafalox]

I commented to manu this idea of having a “beta testing” team of players and he had one in the past that wasn’t really good at finding bugs and stuff but I still think it could be a good idea when it comes to gameplay changes

[Ruze]

Well written Ambience. As an online playerbase I'm sure it can be daunting to interact with us all on this forum, especially with the frequency and care that Manu does, so big kudos to him. Our feedback comes from a place of passion, love, and care for this game that we all enjoy so much. I can tell we all want this game to be the best that it can be. I think the idea of a playtest group or a players' council is awesome and totally support it.

[manutoo]

Ambience,
as Jirafalox pointed out, I already tried that with TE2013, back when updating the game was much easier and faster for me.

It's not that the beta testers weren't good at finding bugs ; it's simply that having hundreds of people testing instead of 4 or 5 is vastly more effective at finding issues. Moreover, a small group of testers will usually have their own ways of playing, which won't cover all possible situations.

Also, regarding the twisted serve change, I honestly didn't imagine it could affect the online balance significantly, so I wouldn't have made a test build for it anyway.

And preparing and managing test builds is actually a substantial amount of extra work for me. It really doesn't fit my development process, so I wouldn't consider doing it except in very exceptional circumstances.

As for polls, don't worry, they're (almost) always informative ; they're useful for gathering feedback and making better-informed decisions, but they're not always necessary. For example, a poll on whether to keep the twisted serve bounce physically unrealistic wouldn't make much sense.

[lewfx]

As a mid-tier player with 500+ hours (and #1 on the most active list for the last 2 months), here’s my take:

The game felt frustrating until replay analysis tools came out (https://replays.knowthew.net/). Reviewing my matches helped me fix my mistakes and see how my "best tennis" looks like.Reached my first Challenger final, beating players I used to lose to. Matches are still super tight and competitive—most of the times it comes down to who handles pressure better

I share these tools with people I beat, but some would rather blame the meta than try to find ways to get better. (Funny there was this guy that found that he was not doing well in long rallies, wanted to be a better server and make that one of his strong points. but he didn't even know about the serve talents)

The reality is that perfect balance across all skill tiers is impossible. Some of the complaints I hear from the community sounds the same as with my time playing League of Legends where there is a "strong meta" and the devs have to be constantly making changes and adjustment so the meta be always changing and you don't have the time and resources that Riot Games has so keep trying your best in this aspect.

Thanks for building such a great game!

P.S.Do you play tennis irl? (The mechanics feel so authentic, never played tennis in my life this is the closest thing I have)
2) Do you think current AI tools will help give Mana Games the horsepower to eventually make Tennis Elbow 5?

[manutoo]

lewfx,
1) I guess I played around 2000 hours on a real court, and I had a decent club level (30/1 in France, equivalent to ~3.0 in USA).

2) As of today, I guess not, but I still didn't try some of the more advanced stuff, and in a couple of years, AI might have made a couple more leaps, so it's hard to predict where we'll stand...

[Ambience]

Ambience,
as Jirafalox pointed out, I already tried that with TE2013, back when updating the game was much easier and faster for me.

It's not that the beta testers weren't good at finding bugs ; it's simply that having hundreds of people testing instead of 4 or 5 is vastly more effective at finding issues. Moreover, a small group of testers will usually have their own ways of playing, which won't cover all possible situations.

Also, regarding the twisted serve change, I honestly didn't imagine it could affect the online balance significantly, so I wouldn't have made a test build for it anyway.

And preparing and managing test builds is actually a substantial amount of extra work for me. It really doesn't fit my development process, so I wouldn't consider doing it except in very exceptional circumstances.

As for polls, don't worry, they're (almost) always informative ; they're useful for gathering feedback and making better-informed decisions, but they're not always necessary. For example, a poll on whether to keep the twisted serve bounce physically unrealistic wouldn't make much sense.

Thanks for the detailed answer, Manu.

I understand that preparing and managing test builds adds extra work, especially for a solo developer, and that it doesn't really fit your development process. As a fellow SWE I completely understand that. The last thing I'd want is to suggest something that creates more work than value for you.

That said, I hope it can remain an option for the future in exceptional cases, especially when you're considering substantial gameplay changes or experimenting with different bounces, animations, or other mechanics where community feedback could be useful.

Maybe there's also a middle ground through modding. For example, Jira created a bounce mod to experiment with different bounce behaviours (If im not mistaken). Perhaps giving modders more options to tweak certain gameplay parameters could make it easier to test ideas without requiring you to create and manage separate test builds. I imagine that could be a more feasible approach with less overhead on your side.

Regarding the twist serve, I understand that you didn't expect the change to have a significant impact on online balance. My point is that, in competitive online play, even seemingly minor changes can have a noticeable effect. Different players are also sensitive to different aspects of gameplay, so something that seems like a small adjustment in isolation can have a much bigger impact in practice.

I also agree that it's impossible to satisfy everyone with every update. Personally, I didn't particularly like the point building system change, but I really liked the AutoPos change. As I've mentioned in another thread, I also think character speed in online Tour could benefit from a nerf or cap. Everyone has different preferences and priorities, and there's no way every change will benefit every player.

I agree with you that polls aren't necessary for every decision, especially when the answer is already clear from a gameplay or realism perspective. I see them more as an additional tool for situations where there's genuine uncertainty or a significant difference in how players experience a mechanic. Combined with the feedback you already receive, they can sometimes help reveal perspectives or potential consequences that might otherwise be overlooked. Ultimately, it's about using them when they provide value, rather than making them a mandatory step for every change.

I know online isn't your main focus, and I respect that. At the same time, it's still an important part of TE4, and gameplay changes inevitably affect its balance and competitive landscape. Even small changes like the twist serve can generate quite a bit of debate, simply because players experience and interpret the game differently.

Ultimately, I hope we can keep this kind of open communication going, because I think it benefits both you and the community. I know managing feedback can be demanding, particularly when players have different expectations, sometimes conflicting opinions, and different ideas about what the game should be. I imagine it can be tricky to navigate all of that while also managing development.

Everything I've brought up is meant in that spirit: to find common ground, not to tell you how to develop your game or add unnecessary work to your plate. I think there's value in finding ways for the community and you to work together, share perspectives, and help TE4 continue to improve and remain enjoyable for everyone. Especially now that you're no longer working on it full-time, I hope we can make the most of the communication we have and find practical ways to contribute where it genuinely helps. And with AI potentially helping you speed up certain parts of development, perhaps there will be new opportunities to explore some of these ideas in the future, without assuming that it automatically makes every part of the process easier.

Thanks again for taking the time to discuss these things with us.

[manutoo]

Ambience,
1) Adding Mod stuff is work too, plus it increases technical debt and requires maintenance.

2) Online is still one of my priorities, though : I've put a lot of time & effort into it over the years.

3) I've been doing this for more than 20 years now, with more than 20'000 messages on the forum (plus plenty on Steam and a ton of emails), and I've always said the game would never have been this good without the hundreds, if not thousands, of users giving their feedback over the years. I don't see any reason for that to stop now. I won't restore the old twisted serve bounce, though.

PS: Pro tip => Next time, ask ChatGPT to avoid repetitions and make the message shorter. It was awfully long to read...

[Ambience]

Noted on all three!

1) Fair, mods = work + maintenance. It was pure brainstorm, not a feature request.

2) Never doubted it. Years of effort on online and still supporting it. Respect.

3) That's exactly what my (awfully long) thank you note was about. And don't worry, I'm not campaigning for the old twist serve bounce, it was just an example.

PS: In my defense, I was aiming for Proust-level prose as a tribute to French literature. And you did say the game owes a lot to user feedback. Some of us just deliver it in bulk. This one's been trimmed though... by Claude, not ChatGPT.