EPTFTF with IdrA (interview) - Page 6
| Forum Index > SC2 General |
HolyArrow
United States7116 Posts
| ||
|
sevia
United States954 Posts
Carmac is a great interviewer as well. | ||
|
Warpath
Canada1242 Posts
| ||
|
FreshNoThyme
United States356 Posts
If results of match-ups are close to 50%, I don't see any problem. If results are skewed and a race is winning 80% of a match-up, there is a problem in balance. Taking the top 50 players of a game using them as examples of balance leads to games becoming UNBALANCED. Looking at large pools of actual results is based on fact. Opinions of skill level is not scientific and will lead to unbalance. Many of you seem to be thinking of scenarios like the following, but using the information incorrectly: To use a hypothetical: Top 30 players 10 of each race TvT 60% TvP 70% TvZ 80% PvP 50% PvZ 50% ZvZ 50% That doesn't mean necessarily mean T is better than the other races. Using such a small sample size does not give accurate results. The 10 players here COULD be better, or the race of the players COULD be better. This small statistic pool does not help in understanding balance. Now, take the following: Top 300,000 players 100,000 of each race TvT 60% TvP 70% TvZ 80% PvP 50% PvZ 50% ZvZ 50% THAT is much more likely to be imbalance. The chances of the 10 T players in the first scenario being, on average, simply BETTER than the Z and P players is MUCH higher than the chances of the 100,000 T players simply being better than the other races. This is statistics 101. Small sample sizes lead to inaccurate results. Dismissing ladder results is to throw away the best tool in understanding game balance. | ||
|
Whole
United States6046 Posts
| ||
|
MonsieurGrimm
Canada2441 Posts
On July 19 2011 12:17 Whole wrote: We don't have a big enough sample size though. The metagame has been changing too much to get 500 high level ZvPs in one "metagame" phase. it's been patched too often for that kind of sample size, as well | ||
|
Frozenserpent
United States143 Posts
I guess one big argument against taking statistics too seriously at this point is that there aren't a good pool of top players to look at. There are only a couple of defining players for each race. A handful of pros that are considered consistently good. Furthermore, the metagame hasn't stabilized yet, and probably won't do so before any further patch changes or expansion occurs. This all makes it very difficult to draw much balance conclusion from statistics. | ||
|
yoshi_yoshi
United States440 Posts
| ||
|
Luppy1
Singapore177 Posts
| ||
|
Gamegene
United States8308 Posts
| ||
|
Goliath0nline
Canada165 Posts
| ||
|
Maffe
Sweden133 Posts
On July 19 2011 06:06 Carmac wrote: Oh no, I've trolled him 11 months ago. I just haven't released it yet. You must spread the trolling wisdom of Uszat. Upload it! Seriously though, great interview with IdrA. A very serious, personal interview has been longing. | ||
|
Omnipresent
United States871 Posts
| ||
|
genius_man16
United States749 Posts
I can't help but like a guy like that.  | ||
|
Dirt McGirt
New Zealand129 Posts
Wonder how that second interview really ended lol! Tarson must have carving up in the club the night before! | ||
|
Bleak
Turkey3059 Posts
On July 19 2011 13:28 Whole wrote: Well, Ladder is designed to give 50% win rate. The MMR system's purpose of existence is to give you an even match and try to make you have a 50% winrate. Also, why balance the game to lower level players? The only imbalance in lower level players is the skill difference. Yep, there are people who win with every kind of strategy in ladder, and in some cases strategy might not even matter, they might be macroing better for example. It is just not a good indicator to look at for balance. They could look for the top pro competition for that. | ||
|
vol_
Australia1608 Posts
| ||
|
Cuddle
Sweden1345 Posts
| ||
|
Shodanss
Greece245 Posts
Good luck with your prodigee as well, hoping to see her interview more players in the future... | ||
|
Night Eyes
433 Posts
Well done on both interviews, on Carmac's side and also the players. | ||
| ||
