You're not wrong... you're just not looking at it in full detail yet. To make the best team you can possibly make, you have to look at so many things. I was of the belief that more than 10 usage points was a flat waste. And since usage adds to a player's salary, you can get more of the other stuff by getting lower usage... load up on the things that do matter (rebounds, efg%, fta, pf, lower tov%, defense).
However, I've come to the point where I dislike only 10 or 10.5 usage points. I've seen several times now how it has crippled my teams... So I've started trying to branch out and get a little more usage... and to do that, you have to have less of something else. It's always a trade-off with this thing. Yeah, you can get more rebounds but at what cost? More turnovers? Less defense? Lower efg%? When you put 24 truly competent owners together in one league, it's the small details that separate them.
I hypothesize that 12.5 usage points is ideal for an up-tempo team. This creates less of the personal possession penalty yet still allows you to maximize the other areas and still get the most out of your high usage guys. I've won with 10.5. I've won with far greater than that. I've seen much salary wasted on those higher usage teams. I've seen fg% and turnover headaches with lower usage ones. I'll just have to see what tale time tells.
RE: Tov% & to48. Tov% is the number the sim uses. However, the sim doesn't provide it for us (major problem that they don't give two ***** about). You can go to pro-ref for many of them, but you have to crunch numbers for anyone who played in an era that they didn't count turnovers. This is the first reason tov% is better than to48.
The second reason is this...
Example: 64-65 Wilt (1787 minute portion) has a to48 of 5.4 - YIKES!! This is terrible. Stay the hell away from this guy!! Right? Let's instead use 90-91 Admiral. He only has a to48 of 4.2. That's much better, right? On the contrary... That Wilt has a tov% of 11.6%, while that Admiral has a tov% of 13.7%.
That means the admiral turns the ball over about 18% more than Wilt (and here's the important part) in the number of times each player touches the ball.
But what is this? You're comparing a top tier (34.7%) usage guy to a 3rd tier (26.6%) one! Well, ain't that some ****?! Let's grab another Wilt then, shall we?
65-66 Wilt (26.5% usage% - same tier as that admiral). This Wilt has a 4.9 to48 average. That's 0.7 more turnovers per 48 minutes than the Admiral. That's still pretty absurd and sounds like the obvious choice between the two is the Admiral.
But wait... this Wilt's tov% is actually only 13.6%. That's 0.1% less than the Admiral... which means that Wilt really turns the ball over at almost exactly the same rate as the Admiral per possession. Since both guys have about the same usage, they'll end up with about the same possessions. Wilt will have more total turnovers because he plays more minutes, but the to/48 will actually be about the same.
How is this, you ask? It's because WIS's (*cough* fabricated *cough*) turnover numbers are based on a time period where the pace of the game was much faster. Everything about this sim is (supposed to be) normalized. Ever wonder why Oscar can't put up numbers like he did IRL? It's because the sim has effectively nerfed the old school players. Their RL stats that were so amazing are inflated because the game was so much faster (faster game = more possessions = more opportunities to do stuff). So by taking percentages instead of raw numbers, you theoretically come up with a much more balanced across the eras approach.
This would be wonderful if WIS did it correctly... but they epically failed at it, so it pretty much makes only a few old players worth using, while the majority are from the 3-point era.
To further enforce that point, let's compare the following two player seasons who you might expect similar numbers from based on many different things and very different ones based on others:
Oscar's triple double season (61-62)
Lebron's record setting assist season (09-10)
(I copied & pasted all of both players' stats, total, per minute & advanced, but WIS forums won't let me post it for some reason... thanks dumb, ******* red X)
Oscar averaged more ppg, rpg & apg than Lebron...
31.9 - 12.5 - 11.4
29.7 - 7.3 - 8.6
Yet in the sim, Lebron will lay waste to Oscar over an 82 game season. Oscar has 8% less usage, 2% less ast% & 2% more rebounds. And he's $600k more expensive. That can be straight up traced to minutes. Check out how the two stack up dollar per minute:
You're paying $2,821.51 for every minute of Oscar.
You're paying $3,234.70 for every minute of Lebron.
If that Oscar had played only the minutes Lebron had played, you're talking about a guy with roughly an $8.3M salary. It'd actually be less than that because it's not a flat calculation when it comes to this stuff... another flaw of WIS... But if it was, then Lebron would cost you $11.6M (and you'd call that a bargain). Oscar's the better "deal" but Lebron flat out gives you more per minute (important in higher cap leagues). But what kind of deal is Oscar?
In Oscar, you're paying over $10M for a third tier usage, slightly above average assist (for a point), good rebounding guard. In Lebron, you're paying less for a top tier usage, above average assist, good rebounding guard. Well, surely there's something that can make up for it. Those 5.5 to48 look terrible! He actually just has 15.4% tov%, so that's actually not bad for a point guard.
Oh.
Lebron has 12.3%.
Oscar has 120 more ftas!!
He won't even come close to getting the same amount as Lebron.
Oh, and Lebron will give you way more threes and a much better efg% than Oscar.
Does this mean that Oscar isn't good in the sim? Not at all. It just means that, especially with the old players, you can't take anybody's stats at face value - per season, game or 48 minutes. It's - literally - all about the advanced stats.
8/10/2011 10:34 PM (edited)