+- International Simulation Football League (https://forums.sim-football.com)
+-- Forum: Community (https://forums.sim-football.com/forumdisplay.php?fid=5)
+--- Forum: Media (https://forums.sim-football.com/forumdisplay.php?fid=37)
+---- Forum: Graded Statistical Analysis (https://forums.sim-football.com/forumdisplay.php?fid=153)
+---- Thread: *S24 W2 Predictions Tallies (/showthread.php?tid=25038)
*S24 W2 Predictions Tallies - goodfortunecoffee - 08-15-2020
Yeah I went ahead and made another one of these. These are strangely addicting to create. W2 was very interesting as predictions were way more spread out. 51 out of a possible 64 combinations were used! Up from 35 in W1. There were also like 4 users who only predicted 4 or 5 games again for some reason. Also the most used combination, dubbed the TomHanks combo because he was the first user to post this, only got 32 usages. Remember that W1 the most frequent combo was used 111 times. I was curious as to why things are a bit more spread out. It might be because the matchups appear closer this time around. Well I actually tested each matchup this time, but before we get into that, here are all the combos and their frequency.
[32]BAL/YKW/PHI/AZ/OCO/NOL(TomHanks) [-173pts]
[19]CHI/COL/SAR/AUS/OCO/NOL(kanako)(1st entry) [-31pts]
[18]BAL/COL/PHI/AZ/OCO/NOL(slothfacekilla) [+7pts]
[17]BAL/YKW/SAR/AZ/OCO/NOL(huskies311) [+74pts]
[16]BAL/COL/PHI/AZ/OCO/HON(qWest) [-135pts]
[13]BAL/YKW/PHI/AZ/SJS/HON(Maglubiyet) [+79pts]
[13]BAL/COL/SAR/AZ/OCO/NOL(Billybolo53) [+254pts]
[12]CHI/COL/PHI/AZ/OCO/NOL(Beefstu409) [-136pts]
[12]BAL/COL/PHI/AZ/SJS/NOL(Matty7478) [+401pts]
[9]BAL/COL/SAR/AUS/OCO/NOL(Crodyman) [+112pts]
[8]CHI/COL/PHI/AUS/OCO/NOL(teddyoojo) [-278pts]
[7]CHI/YKW/PHI/AZ/OCO/NOL(thecanadiancanuck) [-316pts]
[6]BAL/YKW/SAR/AZ/SJS/NOL(Kotasa) [+468pts]
[5]BAL/YKW/PHI/AZ/SJS/NOL(ValorX77) [+221pts]
[5]BAL/COL/PHI/AZ/SJS/HON(Tylus) [+259pts]
[5]BAL/COL/SAR/AUS/OCO/HON(Arvot) [-210pts]
[5]BAL/YKW/PHI/AZ/OCO/HON(gucci) [-315pts]
[5]BAL/YKW/PHI/AUS/OCO/NOL(Philliesphan96) [-315pts]
[4]BAL/YKW/SAR/AUS/SJS/NOL(slate) [+326pts]
[4]BAL/COL/SAR/AZ/OCO/HON(Punter715) [+112pts]
[3]CHI/COL/SAR/AZ/OCO/HON(Arkz) [-31pts]
[3]CHI/COL/SAR/AZ/OCO/NOL(Jimi64) [+111pts]
[3]CHI/COL/PHI/AZ/OCO/HON(kingsleigher) [-278pts]
[3]BAL/COL/PHI/AUS/OCO/NOL(TheCC) [-135pts]
[3]BAL/COL/PHI/AUS/OCO/HON(snazzlepuss22) [-277pts]
[3]BAL/COL/PHI/AUS/SJS/NOL(IHateBobNutting) [+259pts]
[3]BAL/YKW/SAR/AZ/OCO/HON(Skyline) [-68pts]
[3]BAL/YKW/SAR/AUS/OCO/NOL(cho1325) [-68pts]
[2]CHI/YKW/SAR/AZ/SJS/HON(ThunderTitan82) [+183pts]
[2]BAL/COL/SAR/AUS/SJS/HON(jdourtled) [+364pts]
[2]BAL/COL/SAR/AUS/SJS/NOL(sakrosankt) [+506pts]
[2]BAL/COL/SAR/AZ/SJS/NOL(Lowlycrib) [+648pts][WINNERS!]
[2]BAL/COL/SAR/AZ/SJS/HON(Bruins10) [+506pts]
[2]BAL/YKW/SAR/AZ/SJS/HON(Cynoblaze) [+326pts]
[2]BAL/YKW/SAR/AUS/OCO/HON(gbfn) [-210pts]
[2]BAL/YKW/PHI/AUS/OCO/HON(Troen) [-457pts]
[2]BAL/YKW/PHI/AUS/SJS/NOL(flyeaglesfly29) [+79pts]
[2]BAL/YKW/PHI/AUS/SJS/HON(AdamS) [-63pts]
[1]CHI/YKW/SAR/AZ/SJS/NOL(hihihi_62) [+325pts]
[1]CHI/YKW/SAR/AZ/OCO/NOL(2Burkeulosis) [-69pts]
[1]CHI/YKW/SAR/AZ/OCO/HON(Doomraider) [-211pts]
[1]CHI/YKW/SAR/AUS/OCO/HON(DNeon) [-353pts]
[1]CHI/YKW/SAR/AUS/OCO/NOL(bjbren17) [-211pts]
[1]CHI/YKW/PHI/AZ/OCO/HON(Toasty) [-458pts]
[1]CHI/YKW/PHI/AZ/SJS/HON(muse316) [-64pts]
[1]CHI/YKW/PHI/AZ/SJS/NOL(ckroyal92) [+78pts]
[1]CHI/YKW/PHI/AUS/OCO/NOL(Dagumpa) [-458pts]
[1]CHI/YKW/PHI/AUS/SJS/HON(Hallmonitor_20) [-206pts]
[1]CHI/COL/PHI/AZ/SJS/NOL(jzajenius36) [+258pts]
[1]CHI/COL/SAR/AUS/OCO/HON(BayleyIsland) [-173pts]
[1]CHI/COL/SAR/AUS/SJS/NOL(AW13) [+363pts]
[1]???/COL/PHI/AZ/OCO/NOL(Scorp) [-36pts]
[1]BAL/COL/???/AUS/OCO/HON(Sermokala) [-177pts]
[1]BAL/COL/SAR/AZ/OCO/???(Bwestfield) [+212pts]
[1]BAL/YKW/???/AZ/OCO/NOL(???forgottowriteitdown) [-73pts]
The 1st number indicates the frequency of each combo, obviously. The user next to each combo is the 1st person to post that combo. Notice kanako was the first user to respond on the predictions thread and received the 2nd most frequency. Although still a ways behind number 1. Also that point total has to do with the little game I made up where I assume each pick is a 100 point bet, and the winners split the losers pot. These are the percentages of each game and how many theoretical points each picked team would net you.
69.8% 192/275 +43pts 
30.2% 83/275 -100pts 
55.4% 153/276 +80pts 
44.6% 123/276 -100pts 
40.5% 111/274 +147pts 
59.5% 163/274 -100pts 
70.7% 195/276 +42pts 
29.3% 81/276 -100pts 
25.4% 70/276 +294pts 
74.6% 206/276 -100pts 
70.5% 194/275 +42pts 
29.5% 81/275 -100pts Not as extreme as last weeks. If I was really feeling ambitious I could go back to each predictions thread and count up every user's points but ehhh I dunno. Anywho this time around I got curious how much these percentages match up to actual sim testing percentages of each game. Keep in mind I used the W2 file to test, so it should be fairly accurate.
58% +11.8 42% -11.8 53.6% +1.8 46.4% -1.8 49.8% -9.3 50.2% +9.3 62.6% +8.1 37.4% -8.1 44.4% -19.0 55.6% +19.0 67.6% +2.9 32.4% -2.9The green and red numbers indicate how much more or less the consensus is on these teams compared to the actual chances they win. I kinda view it as if the team is green, they are kinda popular. Or overvalued. Or overrated? Maybe Overrated vs. Underrated might be the best way to put it. For instance way more people picked the Otters against the Sabercats then there should've been. Sabercats are testing better than you probably think. So then I backtracked and did Week 1 as well.
Actual(Predictions)
74% (92.1%) +18.1 26% (7.9%) -18.1 49% (38.5%) -10.5 51% (61.5%) +10.5 75.8% (82.5%)+6.7 24.2% (17.5%) -6.7 20% (8.6%) -11.4 80% (91.4%) +11.4 67.2%(30.8%) -36.4 32.8%(69.2%) +36.4 (!!) 24.2% (7.2%) -17.0 75.8% (92.8%) +17.0Look at that huge disparity with the NOLA vs OCO matchup! Really interesting how the prediction percentages were much more extreme in W1 than in W2. 3 matchups went over 90% one way. But anyways. So let's combine both weeks.
+55.4 +19.9 +19.8 +19.5 +8.9 +5.1 -2.6 -12.3 -14.3 -29.9 -33.5 -36.0 Keep in mind from top to bottom this is overrated to underrated. Maybe I should make positive underrated instead? I dunno. Anyway these totals are already behind 2 season weeks. I might get to work on W3 & W4 if I feel up to it, though I may fall behind. I also might ditch the imaginary point metric towards the top, since I think the overrated/underrated metric is much more interesting.