*Correlation of various stats to wins - Printable Version +- [DEV] ISFL Forums (http://dev.sim-football.com/forums) +-- Forum: Community (http://dev.sim-football.com/forums/forumdisplay.php?fid=5) +--- Forum: Media (http://dev.sim-football.com/forums/forumdisplay.php?fid=37) +---- Forum: Graded Statistical Analysis (http://dev.sim-football.com/forums/forumdisplay.php?fid=153) +---- Thread: *Correlation of various stats to wins (/showthread.php?tid=4280)

*Correlation of various stats to wins - JuOSu - 09-07-2017 I did some quick calculations using team stats from season 2 and checking if they correlated to wins. Some of these are not surprising I'd say. I used a Pearson Correlation Coefficient Calculator online. Strong correlation: Total TDs scored (duh): R=0.8191 Sacks: R=0.7931 Passing TDs: R=0.7791 Defensive TDs: R= 0.7413 Moderate correlation: FG%: R=0.6878 Passer Rating: R=0.5011 Tackles for Loss: R=0.4805 Passing Yards: R=0.4735 Weak correlation: Offensive Line Sacks allowed: R=-0.4401 Fumbles Recovered (by defense): R=0.4325 Total Sacks allowed: R=0.4256 Rushing Yards: R=0.4222 Rushing Average: R=0.402 Pancakes: R=0.3955 Sack Yards allowed: R=-0.3392 Turnover forced: R=0.3158 Punt Average: R=0.3142 Basically non-existent correlation: Tackles: R=0.2262 Rushing TDs: R=0.1832 Receiving Average per Catch: R=0.1691 Interceptions (by defense): R=0.1672 Punt Return Average: R=0.1407 Passing Interceptions: R=-0.06 Pass Deflections by defense: R=0.0364 Kick Return Average: R=0.0071 Safeties (by defense): R=0 I think the main interesting thing is that Sacks are so high up, while Interceptions basically didn't correlate to losses at all. A complete turnover ratio would be interesting but I couldn't find those statistics very fast. I also think forcing turnovers in general ought to be more important but again, with proper context of turnover ratio maybe that would adjust. It obviously makes sense that any TD scored makes you win more games. Duh. Hope you enjoyed this! *Correlation of various stats to wins - kckolbe - 09-07-2017 Such a shame we couldn't draft you. I found it really surprising that passing TDs were a strong correlation, and rushing TDs so weak. I guess Omar Wright and King Bronko really messed things up. They led the league in rushing and passing TDs, respectively, only Wright played for a 5-9 team and Bronko for a 12-2 team. Similarly, the Liberty led the league in interceptions caught, which is likely why that correlation was weaker than expected. That explains why turnovers forced in general was higher. Next year, when Liberty gets more wins, I'll expect closer to .4 *Correlation of various stats to wins - 37thchamber - 09-07-2017 Kinda weird that rushing TDs are so much lower than any other type of TD Generally speaking, to get a rushing TD, you'd have to get closer to the endzone, no? That would suggest more total yardage, which I'd assume correlates fairly well to wins. Then I thought maybe it's just that a good passing game in general correlates more strongly with wins (probably because teams tend to pass for twice as much yards as they rush for) but if thats the case, I'd have expected interceptions -- being directly counter to an opponent gaining pass yards/TDs; like sacks -- would correlate more strongly with ... not losing. Which in this case would mean winning. Interesting stuff anyway EDIT: ah nvm. kckolbe pointed out something I completely overlooked, lol *Correlation of various stats to wins - kcheng686 - 09-07-2017 Pretty cool, but the sample size for this might be too small