shoes22
Hall-of-Famer
- Joined
- Dec 4, 2012
- Messages
- 28,748
- Reaction score
- 39,806
- Points
- 148
Alright folks, now that the 1st round is in the books it's time to hold my feet to the fire and see how well my model has done so far. From the initial post on April 18th the model correctly identified the overall winner in 7 out of 8 playoff series. The one it got wrong was the Trailblazers advancing, but accounting for the Chris Paul and Blake Griffin injuries would make it 8 for 8.
It predicted the exact outcome in 4 of 8 playoff series. (Damn you Charlotte for not winning Game 6.) In the 4 series it got wrong, it first predicted the correct outcomes of Raptors in 7 after Game 1, Heat in 7 after Game 3, Trailblazers in 6 after Game 4, and Hawks in 6 after Game 5.
Overall the accuracy was not too impressive, but the real strength of the model is less predicting overall series (which is the result of multiplying odds), but moreso predicting individual games. Through that lens it does much better. The model correctly identified the winner in 39 out of the 44 first round games. The ones it got wrong were Indiana beating Toronto in game 1, Dallas beating OKC in game 2, Houston beating Golden State in game 3, Charlotte beating Miami in game 5, and Miami beating Charlotte in game 6.
Now that I've sufficiently bored you with all that, here's how the second round looks after last night:
Golden State Warriors (1) vs. Portland Trailblazers (5)
Warriors in 4: 21.95% (+7.14%)
Warriors in 5: 30.73% (+4.86%)
Warriors in 6: 18.48% (-0.49%)
Warriors in 7: 15.09% (-3.13%)
Trailblazers in 7: 6.47% (-1.34%)
Trailblazers in 6: 5.53% (-3.07%)
Trailblazers in 5: 1.74% (-2.15%)
Trailblazers in 4: 0.0% (-1.82%)
Game 2 Odds:
Warriors win 70%/Trailblazers win 30%
Overall Odds:
Warriors win series: 86.26% likely (+8.38%)
Trailblazers win series: 13.74% likely (-8.38%)
And the series you've been waiting for...
Toronto Raptors (2) vs. Miami Heat (3)
Raptors in 4: 4.41%
Raptors in 5: 13.36%
Raptors in 6: 11.31%
Raptors in 7: 19.35%
Heat in 7: 12.90%
Heat in 6: 20.89%
Heat in 5: 11.02%
Heat in 4: 6.76%
Game 1 Odds:
Raptors win 60%/Heat win 40%
Overall Odds:
Raptors win series: 48.42% likely
Heat win series: 51.58% likely
It predicted the exact outcome in 4 of 8 playoff series. (Damn you Charlotte for not winning Game 6.) In the 4 series it got wrong, it first predicted the correct outcomes of Raptors in 7 after Game 1, Heat in 7 after Game 3, Trailblazers in 6 after Game 4, and Hawks in 6 after Game 5.
Overall the accuracy was not too impressive, but the real strength of the model is less predicting overall series (which is the result of multiplying odds), but moreso predicting individual games. Through that lens it does much better. The model correctly identified the winner in 39 out of the 44 first round games. The ones it got wrong were Indiana beating Toronto in game 1, Dallas beating OKC in game 2, Houston beating Golden State in game 3, Charlotte beating Miami in game 5, and Miami beating Charlotte in game 6.
Now that I've sufficiently bored you with all that, here's how the second round looks after last night:
Golden State Warriors (1) vs. Portland Trailblazers (5)
Warriors in 4: 21.95% (+7.14%)
Warriors in 5: 30.73% (+4.86%)
Warriors in 6: 18.48% (-0.49%)
Warriors in 7: 15.09% (-3.13%)
Trailblazers in 7: 6.47% (-1.34%)
Trailblazers in 6: 5.53% (-3.07%)
Trailblazers in 5: 1.74% (-2.15%)
Game 2 Odds:
Warriors win 70%/Trailblazers win 30%
Overall Odds:
Warriors win series: 86.26% likely (+8.38%)
Trailblazers win series: 13.74% likely (-8.38%)
And the series you've been waiting for...
Toronto Raptors (2) vs. Miami Heat (3)
Raptors in 4: 4.41%
Raptors in 5: 13.36%
Raptors in 6: 11.31%
Raptors in 7: 19.35%
Heat in 7: 12.90%
Heat in 6: 20.89%
Heat in 5: 11.02%
Heat in 4: 6.76%
Game 1 Odds:
Raptors win 60%/Heat win 40%
Overall Odds:
Raptors win series: 48.42% likely
Heat win series: 51.58% likely