Post by eric on Jan 14, 2016 15:29:30 GMT -6
Last year the 6 seed Knicks met the 5 seed Suns in the TMBSL Finals ® Presented by Mike's Apple Ale ©. This was shocking for many reasons, but it was the seeds that stuck out to me.
The NBA went to a 16 team playoff bracket in 1984 and has played 32 postseasons in it. (Note that the NBA has also expanded from 23 to 30 teams in that time, which makes upsets more likely by putting better teams in the eight seed, for example last year's 45 win Pelicans compared to 1986's 30 win Bulls.) In those 32 postseasons, the teams with the best records in their respective brackets have reached the Finals in 10. In another 15 of those years, one Conference's number one seed reached the Finals. In all of the remaining 7, at least one team with the second best record (usually but not always the number two seed) made it.
This is dramatically different from our TMBSL, but these are small sample sizes so I decided to run a thousand simulations: take the 2005 simulated NBA, run the playoffs in the software 500 times with injuries turned off, run the playoffs 500 times with injuries turned on. Here are the results...
The top row refers to the 1 seed from each Conference making it, then only one, then at least a two seed, then only bottom half teams. The columns are real life, our league, 500 sims without injuries, 500 sims with injuries. Obviously adding injuries makes it harder for the best teams because they have the most to lose, but there's something else in the software that makes upsets much, much, much more likely. We've established that home court advantage is actually slightly stronger than IRL, our pace tends to be way higher than IRL, and losing teams in the software don't play the foul game to give themselves a chance, so my guess is that there is a familiarity/coaching/extra gear effect IRL not reflected in the software.
Or the RL playoffs are just rigged. Either way, #1 seeds going down in our playoffs should be viewed with much less disdain than in real life.
The NBA went to a 16 team playoff bracket in 1984 and has played 32 postseasons in it. (Note that the NBA has also expanded from 23 to 30 teams in that time, which makes upsets more likely by putting better teams in the eight seed, for example last year's 45 win Pelicans compared to 1986's 30 win Bulls.) In those 32 postseasons, the teams with the best records in their respective brackets have reached the Finals in 10. In another 15 of those years, one Conference's number one seed reached the Finals. In all of the remaining 7, at least one team with the second best record (usually but not always the number two seed) made it.
This is dramatically different from our TMBSL, but these are small sample sizes so I decided to run a thousand simulations: take the 2005 simulated NBA, run the playoffs in the software 500 times with injuries turned off, run the playoffs 500 times with injuries turned on. Here are the results...
seeds IRL TMBSL soft hard
11 31% 7% 9% 7%
1 78% 67% 59% 56%
2 100% 78% 88% 87%
bottom 0% 4% 2% 3%The top row refers to the 1 seed from each Conference making it, then only one, then at least a two seed, then only bottom half teams. The columns are real life, our league, 500 sims without injuries, 500 sims with injuries. Obviously adding injuries makes it harder for the best teams because they have the most to lose, but there's something else in the software that makes upsets much, much, much more likely. We've established that home court advantage is actually slightly stronger than IRL, our pace tends to be way higher than IRL, and losing teams in the software don't play the foul game to give themselves a chance, so my guess is that there is a familiarity/coaching/extra gear effect IRL not reflected in the software.
Or the RL playoffs are just rigged. Either way, #1 seeds going down in our playoffs should be viewed with much less disdain than in real life.
