NHL Handicapping
We are about 2 weeks away from the start of the NHL season, which is one of my favorite sports. Some guys like to use the Corsi and some like the Fenwick. And puck heads will argue until the death about which is more important. But they both can be used in our handicapping to help us get an advantage.
Obviously this could be a lengthy discussion, but the short version is to use a team's Corsi against opponents with weak defenders who don't block a lot of shots, and use a team's Fenwick against opponents with defenders who do block a lot of shots. So team A with good defenders is playing team B with lousy defenders. We would use team A's Corsi and team B's Fenwick to determine the expected shot attempts for each team.
But will those shots go in the net? Once the shot is taken and it gets through the defenders, we need to know the skill level of the goaltender. One category I like to use is a goaltenders current save percentage versus his lifetime save percentage. If the current percentage is too far from his lifetime, there will be regression back to his lifetime percentage. So a lifetime 91.5% and a current 94% tells us to expect regression towards the 91.5. His team would likely be overpriced because the public thinks the guy is superman. But a current 88% would also mean regression to 91.5, and the public would be lining up to bet against the team with a blind goaltender. You can see how we can get value in the money-line by avoiding recency bias. FYI - the first year of the Vegas Knights, Fleury's save percentage was 3% higher than his lifetime, which is why a bunch of castoffs were so successful.
Then there's the luck factor. Hockey is probably the luckiest of all the sports. Think of 6 guys in front of the net fighting for a bouncing or rolling puck. It seems like one team scores every time, and another team can't buy a goal. I use the PDO to identify how lucky a team is. Any score between 98 and 102 is neutral. Any score outside that range will expect regression back into the range. So a team with a 104 has been extremely lucky and that won't continue. But a team with a 96 has been extremely unlucky and that also won't continue. I have software that charts the daily PDO of each team so I can see when team's are outside their normal range. It's similar to reading a stock chart.
Travel schedule and injuries are obviously important as it's a long season. An algorithm deciphers all these factors and gives me an estimated percentage chance of each team winning. Convert that to a money-line and compare to the posted lines, and bet the games with a 3% advantage for home teams or 4% for road teams.
My goodness Eppley, that hurts my brain. Just give me a free winner already!
Champions League:
Liverpool vs Napoli Draw +260.
Obviously this could be a lengthy discussion, but the short version is to use a team's Corsi against opponents with weak defenders who don't block a lot of shots, and use a team's Fenwick against opponents with defenders who do block a lot of shots. So team A with good defenders is playing team B with lousy defenders. We would use team A's Corsi and team B's Fenwick to determine the expected shot attempts for each team.
But will those shots go in the net? Once the shot is taken and it gets through the defenders, we need to know the skill level of the goaltender. One category I like to use is a goaltenders current save percentage versus his lifetime save percentage. If the current percentage is too far from his lifetime, there will be regression back to his lifetime percentage. So a lifetime 91.5% and a current 94% tells us to expect regression towards the 91.5. His team would likely be overpriced because the public thinks the guy is superman. But a current 88% would also mean regression to 91.5, and the public would be lining up to bet against the team with a blind goaltender. You can see how we can get value in the money-line by avoiding recency bias. FYI - the first year of the Vegas Knights, Fleury's save percentage was 3% higher than his lifetime, which is why a bunch of castoffs were so successful.
Then there's the luck factor. Hockey is probably the luckiest of all the sports. Think of 6 guys in front of the net fighting for a bouncing or rolling puck. It seems like one team scores every time, and another team can't buy a goal. I use the PDO to identify how lucky a team is. Any score between 98 and 102 is neutral. Any score outside that range will expect regression back into the range. So a team with a 104 has been extremely lucky and that won't continue. But a team with a 96 has been extremely unlucky and that also won't continue. I have software that charts the daily PDO of each team so I can see when team's are outside their normal range. It's similar to reading a stock chart.
Travel schedule and injuries are obviously important as it's a long season. An algorithm deciphers all these factors and gives me an estimated percentage chance of each team winning. Convert that to a money-line and compare to the posted lines, and bet the games with a 3% advantage for home teams or 4% for road teams.
My goodness Eppley, that hurts my brain. Just give me a free winner already!
Champions League:
Liverpool vs Napoli Draw +260.
Comments
Post a Comment