Determining your Best Lineup Combinations
Ever wonder which lineup combinations are your most effective? Maybe you need a defensive stop, or a score in the late game. How do you tell which group of five players gives you the greatest advantage?
As a part of your basketball team’s stats in Hoopsalytics, you have the option to see individual player stats, or stats for every lineup combination for one or more games. In Offensive, Defensive or Team stats, click on the Lineups option.
You’ll see the same stats as you would for individual players, but for every combination of five players. Here’s a portion of the lineup data for my recent Varsity High School team:
You can click on any column header to sort on the values. For example, clicking on Off. PPP (Offensive Points Per Possession) will order this report by the best to worst offensive lineups.
I like the Net PPP and time-adjusted +/- as ways to determine my best lineups, but there are other stats you may value more as a coach. You can sort on just about anything.
Advanced Filtering
Unlike other scoring systems, Hoopsalytics lets you filter on individual players, or specific parts of the game. When you see the Lineups stats, click on Include/Exclude Players to get more options. You can show lineups which only include one or more players – this lets you see which other players best work with a particular player. You can also exclude all lineups with a particular player – for example, if one of your weaker players only gets garbage time minutes, you can remove his or her lineups from the list.
You can also filter on specific parts of the game. For example, you can move the slider to show just lineups in the 4th quarter.
This screenshot has an example – we’ve isolated to just the 4th quarter, included lineups with only Ryan, and excluded lineups with Bella and Annika:
Once you have five or six games scored, you should have enough data to accurately determine your best lineups and player combinations. But as a caveat, you should ignore lineups with little time playing together, as the small sample size may skew your results.
Do you have any novel uses for lineup data? Please share in the comments!