BuzzerBeater Forums

I did not want to start another thread on this, but so far I have been surprised with the lack of Canadian contributions to my cap testing research (RIP and The Mogul are the only Canadians who contacted me). See: (155261.1). Ah yes, of course we will be playing each other in the coming days or seasons, so I could understand if that was the reason why. On the other hand, maybe many of you just do not get a chance to read BB Global very often.

In any case, with about 30+ data points, this is what I can report so far: my assumptions are holding and the model fits really well (very high r-squared). But again, it is really tough to say at this moment if that is because of the lack of data or if I am really on to something.

This is what the model is currently saying:

-Position has no impact on capping
-The model is exponential, similar to the salary formulas, but the coefficients are quite different. This means that the higher the skill, the faster you approach the cap per skill-up.
-Jump shot, jump range, inside defense and rebounding all contribute significantly to capping (and at similar levels)
-Passing and inside shot contribute in a very minor way
-All other skills have zero impact on capping.

While some of this is what I expected, I still find it surprising, for example, that outside defense would have zero contribution. I suspect this is because the model is still quite flawed. Also, while the model is working well on the data for allstar potential, it does really poorly (from my point of view) on any other potential.

Of course, this is a symptom of the data. Most of the players I am getting have allstar potential, so that's why the model is quite good there. Also, I am tending to get more PFs and C, so this is likely the reason why guard skills are not getting represented in the model.

If anyone out there can help with data, I would really appreciate it. Of course I would love to get some variety on the data, but anything helps. If nothing else I can really get a really accurate picutre of allstar Cs. ;-)

Do you have the stats of this dude? (2402564)
Pure insideshot created that salary?

Do you have the stats of this dude? (2402564)
Pure insideshot created that salary?

Thanks for the comments.

I don't have his skills but I see he has bounced around a lot. As such, it would be hard to get a read on where he capped or even if he is capped. I will contact his owner, though, to see what he has to say.

In any case, if he is a pure inside shot monster, it does seem to confirm at least part of what my (preliminary) model is saying. Although it does seem weird. Why would outside shooting impact the cap so much, while inside shooting impacts it so little?

Unfortunately I have only capped out all-stars and have little to add. In fact, most of this was done back in January 2009 (35905.52). I have been wasting my time with my sf's (with one replaced by iruleta/kordalewski) ever since.

Just want to say that this guy(2402564) existed before the potential system came out, which means that the system may not work on him as he had been improved before it. everyone without potential before season 5 were not limited to improve. So I think all the guys with default all-star potential should not be referred to the potential system.

So I think all the guys with default all-star potential should not be referred to the potential system.

I was under the impression that only about 2-3 players were over the allstar cap when the system was introduced. I think Romero was one of them. This guy may be another. But unless you (or someone else) knows otherwise, I will just make sure to not use one of those players. ;-)

Also, I am only using data when the person in question can give me some background on the player's training. So like I said, this almost certainly eliminates the $250K allstar example above since he has bounced around so much. I am not taking players directly off the transfer list, for example. While this system is not perfect, I think it is probably as good as any.

If I start to get a lot of data, I can be even more choosy about what I put in the model. But right now I don't have that option.

Last edited by HeadPaperPusher at 8/30/2010 1:54:16 PM

Unfortunately I have only capped out all-stars and have little to add. In fact, most of this was done back in January 2009 (35905.52). I have been wasting my time with my sf's (with one replaced by iruleta/kordalewski) ever since.

Thanks for that thread, I am a late joiner to Superfly's fed and did not know about it. It seems there are quite a few examples there that I do not have. I will plug them in and see what happens to the model.

I think I am close to capping a star guard but I won't know for sure if he's done until early next season at the soonest. (I suspect he's close but when I switch over to two position JS training next season I will know for sure if he doesn't pop in JS within a reasonable length of time)
I'll let you know when I think he's hit the wall. I suspect Milot is at least two full seasons from hitting his cap, but I could be wrong. We'll see if I am even willing to push his salary that high.

I assume you are talking about Guanglun? He must be getting pretty close, based on his DMI (DMI is a rough estimator of current salary when game shape is proficient). If you want to get him as many pops as possible, I suggest mixing handling, passing and outside d training. At least that's what my analysis says so far. But like I said, it is very preliminary and most likely hokum.

By the way, you can ask The Mogul about Jean-Francis Saint-Martin (7419597). I am pretty sure he is capped at his current skill set. It can give you an idea what a star guard caps at.

A couple of questions. Did you ever take Neascu or Vyberal to their cap? If so, do you remember where they were when they capped.

Also, for Marcum and Oberortner, are the skills listed there the skills when they capped (to the best of your knowledge). Or did you give them a bit of training after their cap?

Thanks again for all the information.

Marcum and Oberortner were capped at that point. Neacsu managed one more OD improvement, but was likely capped at the time of the Jan 09 post. I was unsure if Vyberal was capped - he reached 15 level js, but was injured a few times right before he stopped receiving training.