What if Tim Duncan Took Dwight's Place?
Posted by Brett on 09/21/2009So, what would happen to Orlando if they traded Dwight Howard for Tim Duncan? What happens to the Knicks if they can replace Al Harrington with King James? What would happen to the Bobcats if you replaced Raja Bell with Allen Iverson? Scratch that last one - it will only disappoint me. Well, you know how I came up with that formula in One of the 3 kinds of lies? Well, now you can use that on everyone in the league: Click here to go to the Player Swap Page.
I kept the same formula and the same stats for the hypothetical stats. You can review the basics on the Swap Page (and the link to the full article is also there). An addition I made with this tool: I do not just calculate what the team's efficiencies will be with their new player on the court, I also estimate how they might impact the overall team efficiencies. How's that, you may ask? Well, the initial example was Emeka Okafor and Tyson Chandler, and Emeka played about twice as many minutes last year as Tyson. That can be a problem depending on the quality of a team's bench - but it would not show up when only looking at on court efficiencies. So, I added a few extra rows - expected overall efficiencies when Tyson takes all of Emeka's minutes, when Tyson plays the same number he did last year - giving significantly more to the bench, and when Tyson plays the mid-point in minutes between his and Emeka's numbers. Basically, it penalizes someone if they are forcing their team into giving a lot of minutes to weaker players.
I am not repeating my caveats from the other post - so, look it over, use it and comment - it is by no means a finished product and I would welcome constructive feedback on it (or suggestions for other tools you might like to see it).
Click here to comment/view comments for this post.
| Blazersedge Ben Says: | you're on the next level brett, nice work. | ||
| Tue Sep 22, 2009 2:53 pm |
| Marc Says: | Not working for me--getting divide by zero errors. Is this populated by 2009-10 stats already, or am I doing something wrong? | ||
| Tue Sep 22, 2009 3:00 pm |
| Brett Says: | Thanks, Ben. Marc, what players are you using that you are getting the error for? It is using stats from last season. |
||
| Tue Sep 22, 2009 3:26 pm |
| cool Says: | Cool, i just substituted Nowitzki with Ben Wallace, the Mavs are going 73:9! | ||
| Tue Sep 22, 2009 5:43 pm |
| Brett Says: | Cool - haha. I need to do some tweaks around increased usage causing a decrease in efficiency. That anomaly is caused by the formula assuming Ben Wallace would increase his usage from 3 possessions to 18, and not see a decrease in his efficiency - for extreme cases like his (or at the other end of the spectrum, Lebron), the equation should only use the team's efficiencies to calculate the offensive efficiency. The same would go for LeBron taking the place of someone - he is not likely to significantly decrease his usage, so the equation should be tilted that way, rather than the current equal split of extra possessions. I'll work on that. | ||
| Tue Sep 22, 2009 6:36 pm |
| Crow Says: | Not sure how the defensive efficiency calculation is being done or how it should be. If you explain the former in english step by step for say S O'Neal for B Wallace I might have a response. I can't follow it right now. |
||
| Wed Sep 23, 2009 3:53 am |
| Brett Says: | Crow, the formula is the same for calculating the defensive efficiency, it just uses the Against numbers in place of the personal offensive numbers. | ||
| Wed Sep 23, 2009 7:44 am |
| Crow Says: | I tried manually running your formula for Shaq in Cleveland instead of Wallace and I got a much higher figure than you show for defensive efficiency so I asked for a step by step run thru to check it and see where I went wrong, if I did. One run thru doesn't seem like too much to ask. Hard for me to use without. | ||
| Wed Sep 23, 2009 11:52 am |
| Brett Says: | Crow: Cleveland's defensive efficiency minus the points Ben gave up, plus the points Shaq surrendered 95.9 - 18.6 + 19.1 + Plus the difference in possessions used against them (17.9 - 15.3) * the average of the remaining cavaliers defensive efficiency + shaq's ((95.9 - 18.6)/(100-17.9) + (19.1/15.3))/2 |
||
| Wed Sep 23, 2009 12:22 pm |
| Crow Says: | it wasn't clear to me that pointsA/100 possesssions was individual than rather team efficency while on the court so that was part of the issue. "Plus the difference in possessions used against them (17.9 - 15.3) * the average of the remaining cavaliers defensive efficiency + shaq's ((95.9 - 18.6)/(100-17.9) + (19.1/15.3))/2" (95.9 - 18.6)/(100-17.9) ends being .94 and meaning what? It is not the the remaining cavaliers defensive efficiency. The formula eliminates Shaq's role in help defense while Ben Wallace's is still there with Shaq on the Cavs. |
||
| Wed Sep 23, 2009 1:10 pm |
| Crow Says: | To change things to get at help defense you could use the difference between defensive points allowed per 100 possessions and position average and the player's defensive adjusted rating available here http://tinyurl.com/lkl7fb and explained here http://sonicscentral.com/apbrmetrics/viewtopic.php?t=2294 |
||
| Wed Sep 23, 2009 1:20 pm |
| Brett Says: | "It is not the remaining cavaliers defensive efficiency". Crow - that is exactly what it is in this use. The team's overall efficiency was 95.9 points per 100 possessions. After subtracting the points Ben allowed and the possessions his man used up, we're left with his teammates allowing 77.3 points in 82.1 possessions - a per possession efficiency of 0.94. As for help defense, no, this does not attempt to calculate for that. |
||
| Wed Sep 23, 2009 1:29 pm |
| Crow Says: | For Shaq in Cleveland I can't see 95.9 as the first building block as is. That is the Cavs with Wallace's help defense. But maybe you could go 95.9 - 18.6 + 19.1 -Wallace's net help defense + Shaq's net help defense + the difference in possessions used against them ((17.9 - 15.3) * (shaq's 19.1 - wallace's 18.6)/100)) |
||
| Wed Sep 23, 2009 1:41 pm |
| Crow Says: | Ok, I see .94 is the Cavs defensive efficiency... for one possession. the average of the remaining cavaliers defensive efficiency + shaq's makes sense to me now. I'll adjust for help defense in my use even if you don't / prefer not to. But thanks for the further explanation. |
||
| Wed Sep 23, 2009 1:48 pm |
| sotis Says: | I tried to swap Travis Outlaw for Hedo Turkoglu and came up all zeroes. All other trades I attempted came up all zeroes as well. Please help. | ||
| Wed Sep 23, 2009 4:16 pm |
| matt Says: | sounds fun but i only get zeros. | ||
| Thu Sep 24, 2009 3:09 am |
| Brett Says: | Matt - let me guess - you're using IE? Please go to http://www.queencityhoops.com/playerSwapPageDebug.php and let me know how many alerts pop-up for you when you click swap (after selecting players). there should be 4, one for each player and team. That will help me with tracking this problem down (I am unable to replicate on any of my computers). | ||
| Thu Sep 24, 2009 9:19 am |
