In the current series between the Boston Celtics and New York Knicks, the Knicks have won twice in Boston, Jayson Tatum is only averaging 18 points per game (26.8 season average), and Spike Lee skipped game 1 for the Met Gala. Important data points for constructing some horrible bets for game 3.
What does the model say?
If the data model could talk, it would probably tell me to stay away from most of the player prop bets for Celtics players. That isn’t to say that the players won’t perform well. It is more of a “IDK WTF they are going to do” message. Nonetheless, I will pull and scrutinize some of the numbers and attempt to hobble together some bets.
Below are the players and predictions that we will consider. The over/under lines and odds at time of writing are in parenthesis.
Jrue Holiday
- 6 assists (3.5, -104)
- 4 rebounds (4.5, +114)
- A combo bet of over 8.5 REB + AST is at +132 odds
Jaylen Brown
- 18 points (21.5, -104)
- 2 3-pointers (1.5, -148)
- 4 assists (3.5, +136)
Al Horford
- 6 rebounds (6.5, +106)
Jalen Brunson
- 23 points (26.5, -115)
- 6 assists (6.5, +104)
Josh Hart
- 25 PTS + REB + AST (26.5, -120)
Karl-Anthony Towns
- 23 points (19.5, +102)
- 10 rebounds (11.5, +106)
If I were to ignore the combos and adjust the predicted numbers to make a parlay bet with a cash out option, the odds would be over +7000 (as of 11:17 AM CT). There is a reason why bets at these odds are referred to as “lottos”. I may do it for fun, but I also keep in mind that $1 can still buy a pack of ramen noodles.
Also, I realize that Jayson Tatum is not included. He could have a 30-point triple-double game. Or a 12-point triple-double game. Or he could have more rebounds than points. Moving on.
Now that I have narrowed down some prop bets, it is time to check more data to be a little less foolish but still foolish, nonetheless.
Second and Third Opinions
Stathead and the Sports Reference websites are my go-to websites for viewing basketball statistics. I also recently started using PropsMadness for the pretty charts.
Time to cherry pick some stats to support my predictions and ignore all the stats that may prove them wrong. I am probably just kidding.
Below is a table from Stathead showing Jrue Holiday stats from his last 5 games playing against the Knicks in New York. In four out of five games, he had at least 4 rebounds. This gives me a little more confidence with a prediction of 4 rebounds.
However, he only had 6 assists in two of those games and in one of those games he was still playing for Milwaukee. There are other variables to consider (e.g. these were regular season games, one game went into overtime, starting lineups, etc.) but the information is still useful.
Below is a chart from PropsMadness. The site will let you get gaga with granularity, but I will keep it relatively simple. This chart displays Holiday’s assist numbers in away games this season (including playoffs if applicable) with the expected active players: Jayson Tatum, Jaylen Brown, Derrick White, and Al Horford.
The data shows that within the games fitting this criterion, Holiday has only had one game with 6 assists. After looking at Holiday’s game log for this season, he has only had 16 games (including one playoff game) where he had at least 6 assists. Holiday to get 6 assists is currently at +460 odds. So, the prediction is either dastardly drunk or benevolently brilliant.
After viewing historical data for the other players, I have narrowed down the possible bets.
- Jrue Holiday over 4 rebounds (-200)
- Jaylen Brown over 4 assists (+130)
- Jalen Brunson over 6 assists (-180)
- Josh Hart over 24.5 PTS + REB + AST (-178)
Of these four bets, Jaylen Brown with 4 assists has the best odds at +130. He has only hit this line twice in the current playoffs. However, in away games during the regular season, he had at least 4 assists in over 75% of those games. So, for a straight bet, I will roll with this.
And Now For Something Completely Wild
Every now and then, the model will come up with a prediction that when compared to historical data or over/under lines, seems wild. Keep in mind that data models don’t predict with 100% accuracy. Even though the model that I use has been conservative for the playoffs, it still has misses. Last night, it predicted Donovan Mitchell for 30 points (he scored 43) and Bennedict Mathurin for 13 points (he scored 23 points). But it also predicted Darius Garland for 20 points (he shot 3-11 and scored 10 points).
Part of sports betting is interpreting the data and trying to weed out the misses to make “good bets”. Whatever that means. Another part is just taking a swing for fun.
The model predicted Payton Pritchard for 19 points. His points line is currently at 8.5 (-111) on FanDuel and 8.5 (-120) on DraftKings. On FanDuel, his odds for 15 points are at +430 and odds for 20 points are at +1300. I would usually skip past this because I am currently into making lower risk straight bets. But today I swing.
