Post by Andrew on May 12, 2020 15:15:58 GMT -8
Some of us have over a decade of experience with FBB while others are brand new to dealing with the somewhat enigmatic sim engine that we all love or will grow to love. This article is aimed at providing insight to the latter group, while perhaps enlightening some of the longer tenured GM’s about what it takes to win. What I’ve done is run correlation between wins and all the major statistical categories, both from the team standpoint and from the opponent standpoint. I feel somewhat comfortable with the validity of these outcomes given we are over 80% of the way through the season, although this is a project I would like to continue into the future. As with nearly all statistics, take this with a grain of salt.
Wins:PPG: .799. At the risk of ruining any surprise here, wins and PPG have the strongest positive correlation of any of the categories I compared. It should be of no surprise that you need to put the ball into the hoop to win games. As I recently said: “The best Defense is a good Offense”.
Wins: RPG .751. Part of the motivation for writing this article and conducting this “analysis” was Play and insane talking about how badly they needed to improve their rebounding, almost to the exclusion of other statistical categories. Rebounding has always been an important part of FBB, but this outcome was surprising to me. Wins:RPG was the second highest correlation and is much closer to the Wins:PPG correlation than I thought it would be. Based off of this somewhat small sample size, it would be prudent to improve your rebounding if you’re looking to compete.
Wins: APG .591. Passing is one of the more scrutinized attributes and with good cause. We’ve seen countless examples of winning teams with guards that have low assists. How important are assists really? Are assists the mostly empty statistic many GM’s believe them to be? I’ve always thought of an assist as a somewhat required outcome from the ball going in the basket and not a way of increasing the scorer’s chance of scoring. This correlation shows that assists are indeed important to winning games. However, this could just be an outcome of higher APG implying higher PPG. I would definitely like to see a larger data set about this correlation going forward because I did not expect this number to be this high.
Wins: SPG .348. Ah yes, now we are starting to get into the most fundamental equation in FBB history. Are your team’s steals + blocks > TO’s? If the answer is yes there is a very good chance (in my anecdotal experience) that you at least have a .500 team. However, are each of these variables equally important? It turns out that the answer to that question is no. Wins:SPG is the weakest positive correlation in this data set, which is another surprise to me. In my eyes I view a steal as an immediate 1:1 counterbalance to a TO but this initial analysis seems to indicate that steals aren’t as important to success and wins as I previously thought.
Wins: BPG .662. The second crucial element of the equation. Everyone knows blocks are important to winning, but it is difficult to quantify how they contribute to wins. What happens after a block? What are the determining factors of the ball going out of play? What is the probability that the ball stays with the offensive team, thus nullifying the block? While we can’t definitely answer these questions without a deep dive into FBB’s unavailable code, what I can say based off of the first 65ish games of the season, is that you need blocks to win. Wins:BPG is the third strongest correlation and a much stronger relationship than Wins:SPG
Wins: TOPG -.636. Our first negative correlation doesn’t really provide us with any meaningful or surprising insight. If you turn the ball over more often, chances are you will not win games. Before I aggregated the data I would have guessed that this number would have been closer to -.75 or -.8. Because of the difference in my expected result vs the actual outcome this is a category I am particularly interested in watching over time to see how much the numbers deviate from this initial calculation.
Wins to Opponents stats
I also ran the numbers for wins against opponents statistics just for fun to see if anything surprising came up. There won’t be as much commentary because you can’t, in any major way, affect how other GM’s form their squads.
Wins: oPPG -.245. Surprisingly weak relationship. You can still win games with opponents putting up big numbers against you. (See: Miami)
Wins: oRPG -.321. Nothing to really say here.
Wins: oAPG -.517. Interestingly strong negative relationship considering how weak Wins: oPPG is. The notable discrepancy between wins:oAPG and wins:oPPG is worth keeping an eye on in the future
Wins: oSPG -.589. Strongest negative relationship from the opponents point of view. Worth noting that your own steals per game aren’t that strongly correlated, but your wins to your opponents steals is the strongest negative correlation.
Wins: oBPG .122. A surprising revelation and almost certainly due to the small sample size. The data indicates that wins and opponents blocking your shots is positively correlated! This will surely change over time.
Wins: oTOPG -.210. Nothing of substance and joy here.
If this gets some positive traction I might do the same thing, except from the tanking teams point of view. Hope you enjoyed the correlations. For fun examples of why correlation does not imply causation please see: www.tylervigen.com/spurious-correlations
Wins:PPG: .799. At the risk of ruining any surprise here, wins and PPG have the strongest positive correlation of any of the categories I compared. It should be of no surprise that you need to put the ball into the hoop to win games. As I recently said: “The best Defense is a good Offense”.
Wins: RPG .751. Part of the motivation for writing this article and conducting this “analysis” was Play and insane talking about how badly they needed to improve their rebounding, almost to the exclusion of other statistical categories. Rebounding has always been an important part of FBB, but this outcome was surprising to me. Wins:RPG was the second highest correlation and is much closer to the Wins:PPG correlation than I thought it would be. Based off of this somewhat small sample size, it would be prudent to improve your rebounding if you’re looking to compete.
Wins: APG .591. Passing is one of the more scrutinized attributes and with good cause. We’ve seen countless examples of winning teams with guards that have low assists. How important are assists really? Are assists the mostly empty statistic many GM’s believe them to be? I’ve always thought of an assist as a somewhat required outcome from the ball going in the basket and not a way of increasing the scorer’s chance of scoring. This correlation shows that assists are indeed important to winning games. However, this could just be an outcome of higher APG implying higher PPG. I would definitely like to see a larger data set about this correlation going forward because I did not expect this number to be this high.
Wins: SPG .348. Ah yes, now we are starting to get into the most fundamental equation in FBB history. Are your team’s steals + blocks > TO’s? If the answer is yes there is a very good chance (in my anecdotal experience) that you at least have a .500 team. However, are each of these variables equally important? It turns out that the answer to that question is no. Wins:SPG is the weakest positive correlation in this data set, which is another surprise to me. In my eyes I view a steal as an immediate 1:1 counterbalance to a TO but this initial analysis seems to indicate that steals aren’t as important to success and wins as I previously thought.
Wins: BPG .662. The second crucial element of the equation. Everyone knows blocks are important to winning, but it is difficult to quantify how they contribute to wins. What happens after a block? What are the determining factors of the ball going out of play? What is the probability that the ball stays with the offensive team, thus nullifying the block? While we can’t definitely answer these questions without a deep dive into FBB’s unavailable code, what I can say based off of the first 65ish games of the season, is that you need blocks to win. Wins:BPG is the third strongest correlation and a much stronger relationship than Wins:SPG
Wins: TOPG -.636. Our first negative correlation doesn’t really provide us with any meaningful or surprising insight. If you turn the ball over more often, chances are you will not win games. Before I aggregated the data I would have guessed that this number would have been closer to -.75 or -.8. Because of the difference in my expected result vs the actual outcome this is a category I am particularly interested in watching over time to see how much the numbers deviate from this initial calculation.
Wins to Opponents stats
I also ran the numbers for wins against opponents statistics just for fun to see if anything surprising came up. There won’t be as much commentary because you can’t, in any major way, affect how other GM’s form their squads.
Wins: oPPG -.245. Surprisingly weak relationship. You can still win games with opponents putting up big numbers against you. (See: Miami)
Wins: oRPG -.321. Nothing to really say here.
Wins: oAPG -.517. Interestingly strong negative relationship considering how weak Wins: oPPG is. The notable discrepancy between wins:oAPG and wins:oPPG is worth keeping an eye on in the future
Wins: oSPG -.589. Strongest negative relationship from the opponents point of view. Worth noting that your own steals per game aren’t that strongly correlated, but your wins to your opponents steals is the strongest negative correlation.
Wins: oBPG .122. A surprising revelation and almost certainly due to the small sample size. The data indicates that wins and opponents blocking your shots is positively correlated! This will surely change over time.
Wins: oTOPG -.210. Nothing of substance and joy here.
If this gets some positive traction I might do the same thing, except from the tanking teams point of view. Hope you enjoyed the correlations. For fun examples of why correlation does not imply causation please see: www.tylervigen.com/spurious-correlations
/cdn.vox-cdn.com/uploads/chorus_image/image/65552195/tyler_parker_5_guys_kings.0.jpg)
