Quote: Originally Posted By monkee on 8/12/2008I still havent heard a solid one for normalizing FG%

Perhaps it was done so that more players from the 50's could be used in the sim.

normalization, bullsh*tization

please please please gentlemen, let's not start splitting hairs...

when you draft in the baseball sim you have to compare real stats to # stats to + stats... nobody knows what to expect anymore some swear the # is what matters others think + is the answer others believe real stats is more relevant so that they now have the performance history to look at which is even more confusing because of all the different ballpark factors and some owners overuse their players to death so that completely throws off whatever #, + or real stats you're trying to figure out...

Baseball at some point needed some kind of normalization because of it's very long history with very different eras such as, deadball, free triples, '90's steroid homerun derby, pitchers who pitched 900 innings and so on.

I don't care for basketball normalization, I wanna know what I draft not guess which option, #, +, % or real is going to be more effective...

Let's stay on the real problems we have here at the moment, defensive ratings and offensive range attribution.

One more thing I'd add, players with extreme low stamina should get way more injuries, 0% should be an automatic injury, the way it is now let's owners abuse it by drafting very low minutes and still get away with it.

You miss my point, tianyi. By using a fixed normalization target, you are doing the equivalent of taking all the players and dumping them on another planet where the conditions are different than any of them actually experienced.

By adjusting to a normalization target based upon the players playing, a league of only current players would play based upon current conditions, and a league of only dead ball era players would play based upon dead ball era conditions. If you have mostly current players and only a couple dead ball era players, then it is the dead ball era players whose stats would see the big adjustment. And vice versa ...

Then you have the What-If situation of what happens if A Rod played at the beginning of the 20th century or if Ty Cobb played today. And if the mix of players is diverse enough, you would still get something like what the fixed average would provide.

Quote: Originally Posted By bball_jones on 8/14/2008Baseball at some point needed some kind of normalization because of it's very long history with very different eras such as, deadball, free triples, '90's steroid homerun derby, pitchers who pitched 900 innings and so on

Wouldn't this be easier if salary was adjusted? A 9 mil Hakeem vs. a 9 mil Elvin Hayes.... Very few would choose Hayes. But if Hayes was 6.5 mil (more likely his worth in terms of sim output) he may be more useable... and you are getting more "real" stats since nothing is adjusted.

Quote: Originally Posted By bball_jones on 8/14/2008when you draft in the baseball sim you have to compare real stats to # stats to + stats... nobody knows what to expect anymore some swear the # is what matters others think + is the answer others believe real stats is more relevant so that they now have the performance history to look at which is even more confusing because of all the different ballpark factors and some owners overuse their players to death so that completely throws off whatever #, + or real stats you're trying to figure out..

If the # number is provided, there is no need to compare the real life stat to the # stat. The sim engine uses the # stat.

ncmusician_7 got the biggest reason. Most of the guys from the 50s and 60s were unusable because of the very low shooting%. The league average FG% ranges from 36.7% in 1951-52 to 49.2% in 1983-84. That's a very significant difference, which is why we decided that normalizing FG% makes sense.

With that said, it's not like some guy that shot 39% is suddenly going to shoot 50%. Generally it'll be within a couple percent either way of what they actually shot. In the SimEngine it depends on the 5 defenders' league average as well since that affects the defenders' FG% allowed. But we still don't use an entirely normalized percentage. It's a weighted combination of the normalized number and the actual number.

Quote: Originally Posted By seble on 8/14/2008ncmusician_7 got the biggest reason. Most of the guys from the 50s and 60s were unusable because of the very low shooting%. The league average FG% ranges from 36.7% in 1951-52 to 49.2% in 1983-84. That's a very significant difference, which is why we decided that normalizing FG% makes sense.

yep - another 'based on nothing', subjective decision from our resident hyopcrite who refuses to make changes to individual player status without the most rigorous proof - and apparently the word of the actual coach isnt enough (dcollins on 88 MJ) - but when it comes to a couple decades worth of players has no qualms about making changes that impact game play in a much more significant way without a single objective criterion in sight

MJ 88-89 100% PG PLEASE

Quote: Originally Posted By ncmusician_7 on 8/14/2008If the # number is provided, there is no need to compare the real life stat to the # stat. The sim engine uses the # stat

From my own personal experience and from what other owners have been debating about in the MLB forums it doesn't seem to be that simple. You can't really expext the sim to give you only # stats, it's more like seble said, somewhere between real stats and normalized stats.

To test the FG+ normalization I simmed 20 games

83-84 pistons vs 53-54 celts and looked at Cousy and Isiah FG%. They were both very similar defensively so that effect should be constant. INRL Cousy shot 38.5% to Isiahs 46.2% but Cousy was 103fg+ to Isiah 94fg+.

In the 20 game Cousy shot 41.6% and Isiah 46.6%. The normalization didn't appear to have a very big impact in this 20 game sim.

20 game is a pretty small sample qistat. Statisticians generally require at least sample size of 30 in a randomized pool, in this case, I would say you won't get "accurate" results unless u sim at least 100 games.

But the problem is, seble already said that FG+ has no effect on the sim, thus how exactly was FG% normalized?

Quote: Originally Posted By seble on 8/12/2008eFG% and FG%+ are just for display. They're a way to evaluate the worth of players.

But we don't simply use FG% either. 2 point shots start with the player's 2pt% and 3 point shots start with 3pt%. But there are a number of modifications (e.g. def positioning, team assists, normalization, etc.) applied before we come up with a final adjusted percentage.

FG%+ is useful in evaluating how a player will fare during the normalization process. As it's been pointed out, it's not a complete picture. Right now, we're leaning toward replacing it with the 2pt%# and 3pt%#, which would be the normalized percentages against a historically average opponent. I think it would make sense for us to add a more detailed explanation of how normalization works in the SimEngine to the Knowledge Base. That would probably help clear up these different FG% numbers

For reference

This post could not be converted. To view the original post's thread, click here.

This post could not be converted. To view the original post's thread, click here.