Djoker wrote:lessthanjake wrote:
I did some calculations for Magic below:Spoiler:
If I did it for just the 1988, 1990, and 1991 data we have from Squared, it comes out to just a +7.52 on-off, with a +9.25 ON.
That said, I noticed today that the on-off estimate I got for Magic in 1991 is pretty different from what you listed in the 1991 RPOY thread (mine is worse). So I may have screwed something up in these Magic calculations? Or maybe you actually have detail of the specific games Squared used for Magic (whether that year or all years)? I vaguely recall seeing game-by-game data for Magic on Squared’s website, but I can’t find it anymore so I couldn’t use it in my calculations. I instead had to just use various assumptions (i.e. (1) that the Lakers’ average MOV in the sampled games was the same as their average MOV during the season; (2) that Magic’s MPG in the sampled games was the same as his MPG in the season as a whole; and (3) that the Lakers’ average number of minutes played per game in the sample was the same as it was in the season as a whole—i.e. that they had an average number of OTs in the sample). For 1991 in particular, I can see assumption #1 skewing things against Magic, because the Lakers had a notably worse record in the sampled games than overall, so I’m probably significantly overestimating how well the Lakers did in the OFF minutes that year, though that may not be true overall in the multi-year samples. Oh, and another assumption I made was that Magic played in every sampled Lakers game. He didn’t miss many games in those years and I suspect Squared was much less likely to sample a Lakers game Magic didn’t play in, but it is nevertheless an assumption that might cause error. That wasn’t a problem when I did the same calculations for Jordan in 1990, since Jordan played every game that season.
So yeah, I’m not super confident in those Magic calculations. Regardless of estimation issues, I definitely agree that Squared’s data improves my view of Magic, though!
And I look forward to your upcoming Jordan thread!
________
EDIT: I notice you have some precise data here: https://forums.realgm.com/boards/viewtopic.php?f=64&t=2402176. Where were you able to get the info for that? Is it stuff that is no longer on Squared’s website or is it there and I’m just not able to find it? Of course, the data in that thread doesn’t include tallies Squared made since then (including 1990), but that info looks similar for Magic too. And I think, even accounting for the data in your post being in per-100-possession terms instead of per-48-minute terms, it’s definitely supportive of Magic’s on-off data looking up there with Jordan/Jokic/Steph.
For Magic, I don't actually know which games were sampled in 1990 and so I don't have his exact numbers including that season. But yea the thread I started is all the available data for 1985/1987/1988/1991 that Squared shared that is not on the website.
By the way, I know exactly which games were sampled for Jordan in 1990.
That is an image of all Jordan games that were not sampled since Squared2020 doesn't (currently) have the footage.
Taking the other 56 games that were sampled, the Bulls are +333 with Jordan in 2201 minutes. Without Jordan they are -182 in 507 minutes. That works out to +7.5 ON -17.8 OFF +25.3 ON-OFF per 48 minutes. With 1990, unlike 1991, the sampled and unsampled games aren't that different in terms of MOV. In the 56 sampled games, the Bulls had a +2.7 MOV and in the 26 unsampled games a +4.5 MOV.
By the way, Squared currently has 130 games left to do for Jordan. 2 in 1985, 42 in 1989, 71 in 1992 (MJ actually played 69 of those) and 17 in 1995. I hope that more Bulls games surface. I am certainly trying to help him find some of those games. So far I got him one game!It's actually tough tracking down rare games that aren't widely available and only owned by a couple of collectors. First you had to find them and then they have to be willing to part with the games for a reasonable price.
For Magic, I don't know the exact number of games left to log but I think there's about 40 games in 1986 and 50 games in 1989 left to do.
That is fantastic info! If we know the specific games for 1990, then I can actually add that info to the OP. Will aim to do that today.
EDIT: Have just added the 1989-1990 Squared data to the OP. I will note that the ON and OFF numbers I got for 1989-1990 were different than what you listed above, but I used the plus-minus and minutes numbers you gave to get them, so I’m assuming the +7.5 and -17.8 numbers you gave were typos. Just wanted to flag in case those were the numbers that were right though.
