Mana Games

[JackEnglandisMyCity]

Because I am bored and insane.

I want to see how playing styles of tennis elbow 4 changes according to the type of playing style of the opponent as well as surface. The only other experiment is this sort I could find on the forum was this one- topic15-32164.php?start=15. Most of the methodology in this experiment is inspired by the one mentioned above but I have put my own wrinkles to it. The following are the parameters of the experiment. If you have any criticisms of the methodology and have ways to improve it, pls do leave a comment. I will only list the changes below-

1. At 4), the experiment I mention was made with every player have 90% with a topspin skill at 50% and with both motivation and doubles spirit at very high. This doesn't make so much sense to me since different attributes probably favor different playstyles. However, Manutoo mention that this would be useful for him at the time for some reason. If he does see this, I would be interested in his reasons. Until then, I will be using the best player of each playstyle. (based on Ultimate Tennis Stats' Goat List which I understand is sort of an arbitrary system, but I needed some way to rank them). Constancy, Cold Blood, Concentration and Double Spirit will be all tuned up to 100 in the Players.ini for balance. Talents were be discarded. All players will be normalized at No.1 ranking. Custom playstyles will not be used.

2. Since men and women tennis have different dynamics, I will be carrying out the experiment for both. Men will have a height and weight 185 and 80 respectively, women will be 175 and 60.

3. I'm curious whether the number of sets favor certain playstyles. So there will be 10 matches of every set (Best of 1, Best of 3, Best of 5 (men only))

4. Quality of the game is set to low because I neither have a monster of a pc nor the patient of a monk.

5. The experiment will be carried out on Build 107

6. For balance reasons, all players will play with 100% form of the day. If one player is above or below 100%, the match will be discarded and another match will take its place.

7. The experiment will be carried on the five most important surfaces (AO, RG, W, USO and Tour Finals).

8. Everything else remains the same to the experiment I mentioned above.

This is the list of players that will be used.


1. Defender- Borg

2. Baseliner- Connors

3. Puncher- Lendl

4. Attacker- Sampras

5. Volleyer- McEnroe

6. Counter- Hewitt

7. Counter Puncher- Djokovic

8. All Rounder- Federer

9. Bulldog- Nadal

Same player stats applied to both genders (not sure if that's a good idea or not.) will be starting it in a few hours. Any suggestions would be appreciated.

[manutoo]

Hello,

a) you should probably wait for tomorrow's update to avoid any issues with the weird slice bounce.

b) having all skills at 90% means the strength or weakness of a situation depends on the differences in the style strategies rather than the player skills.

c) 10 matches are probably not enough to conclude for sure, due to the randomness of the results, but it should show some outlines

[JackEnglandisMyCity]

b) I see but would a balanced skill set not favour more universal play styles such as an all court attacker or an all rounder than specialists playstyles like a power baseliner or volleyer? Also wouldn't setting the top spin 50% favour more mid surface styles like counter puncher than those that rely on flat shots like counters or those that rely on spinny shots like defenders?

c) What amount would you recommend? 25? 50? 100?

[manutoo]

b) It's possible, but not sure, that it'd advantage the all-rounders, but it won't be possible to find something more balanced for your tests than having 90% (or 80%) everywhere.
Same for the topspin : it changes a lot of things and having different topspin skills will mean that you won't be only testing the style strategy and but also the topspin effect.
You can totally tune the skills & topspins for the different characters if you want, though, but just be aware that it'll make the results non-exploitable to tune the player styles.

c) Ideally, you'll make a stat after each match and would stop once you get a convergence over a few matches.
Just yesterday, I did ~30 sets per test to tune a new talent to see if it was overpowered or not. With 30 sets, I couldn't tell for sure if it had no effect or a bit too strong effect, or something in between. (so in the end I just trusted my calculations and future user feedback )
I guess after 30 matches you should get a decent idea of who is the strongest player. (ie: with a certainty of 98+%)