Draft Analyzer

[Guest]

For what it's worth, I've observed these Stelmagic numbers to be especially wacky for LBs. They don't seem to have a great handle on LBs. I've had to rely on my own intuition there.

I mentioned this to somebody in the CFL and they agreed; I forget who it was, though.

It seems to value run D a LOT.

[Guest]

So, looking at this, you can see DA said he was 1.21 pick. It also said 28% chance to be Very Good, and 51% chance to be Good. He's none of that :)

His rating is way, way down compared to those around him.

How did you sort it? Slot by round and let it sort itself? If so, slot by round isn't meant to be an actual ranking but where DA's calculations think the player will go. Something like that.

I don't like slot by round and prefer to not use it.

[IrishGuy65]

Here's my ratings of the ILBs... I lied, by the way, he was #5, so he was a top 5 ILB. Like Bergeron, Colin Washington (the guy I would have taken), dropped quite a bit also. Bergeron's 3 best traits were RD, M2M, and ST. Washington's only weakness is his RD, which dropped quite a bit also.

None of them were really very good, and I should have just went a different way with that pick, or tried to trade for a better ILB.

[IrishGuy65]

For what it's worth, I've observed these Stelmagic numbers to be especially wacky for LBs. They don't seem to have a great handle on LBs. I've had to rely on my own intuition there.

I mentioned this to somebody in the CFL and they agreed; I forget who it was, though.

I've never gone by other's information before, always just worked my own way. Sometimes it works out, sometimes it doesn't, because I'm trying too hard. but at least it's my own mistakes, and not someone else's.

That why this particular case bothers me so much... it wasn't my own mistake in rating, it was someone else's mistake that I went along with.

[Aston]

It seems to value run D a LOT.

Hm. I haven't quite put my finger on it. But it definitely seems to be extremely optimistic about pretty average LBs sometimes.

In most other cases, the DA numbers kind of confirm and reinforce what you might expect. It just doesn't seem to be the case with LBs, so some caution is advised.

[Guest]

Also, what all you guys are showing is the evaluation based on YOUR input, not the Draft Analyzer input. If you go to edit, the choose slot by rounds, you will see that it orders them with its own settings... THIS is the DA I'm talking about. I can change the settings around to make whatever I want. As I said, with my personal settings, Bergeron wasn't a top 5 ILB. With the DA settings, he was supposed to be the 21st pick in the draft.

I made the mistake of going with the DA settings. And, after talking with Greg, the settings are based strongly on FOF2k7 settings, which include masked pairs and static bars. As I've been preaching almost since FOF7 came out... these don't have the meaning they used to have. But I still have people disagree with me.

Ah, sorry for the confusion. I've never actually used the slot by rounds option before. Heck I don't even export a draft order from the Analyzer. I use it as mostly as just a tool to see how someone else would rate a player when I'm stumped and cannot decide between two or more guys.

OK, so

1) so open the 2038 draft_personel.csv in the Draft Analyzer, I don't recall ever changing any other settings in the DA, but just in case...

2) open "edit weights" and hit "Default" and "OK", I'm not sure if there are any other options I might have changed int he past, I suppose I could uninstall the DA, find anything the uninstaller leaves behind and delete those (just in case it leaves behind any config files) and reinstall from scratch. But that sounds like a lot of work and I am lazy. "Default" button it is.

3) don't hit anything else besides "sort by rounds" and

4) find Mr. Bergeron. There he is 8rd pick, 3rd round. 72nd player overall.



His Comb. and Rate numbers are a bit higher, so I must have changed something at some point. I wonder if I saved those changes?. Either way, doesn't look to have made that much of a difference for how the DA interprets the draft_personel file as seen by my coaching staff. Which is to say radically different from how other people's coaching staff saw him. Which is the reason for the pop-up "don't share this with other people" warning when you export the draft_personnel file from the game.

I'm assuming there is still some randomness or masking in how your coaches see players, but it is hopefully reduced or offset by them being good at... something. Other people (garion) would be better suited for explaining which bars on which coaches are the important ones for your coaches. I try not to think about the whole coaching system and staff draft as much as possible. It makes me grumpy.

*edit* Turns out I did save those changes to the DA I had made, with them in place "Sort by round" has him going slightly higher, 3rd pick 3rd round. I'm going to see how far I'd have to tweak the settings to get him up into the first round.

Still seems to me that the primary reason for the radical differences we see in certain players from one owner to another are a result of differences in the draft_personnel file created by the game, and the most likely reason for those differences are our different coaching staffs. I agree with IrishGuy, the reason for our different results is different INPUT. But I don't see that as weakness or failure on the part of the Draft Analyzer. Garbage in, garbage out. Maybe next season I'll sack my entire coaching staff try to replace them in incompetent yahoos and see how many first round busts I can find.

[IrishGuy65]

I thought the same thing, Rando. He wasn't that good, IMO. But, I guess the weights I use affect the slotting as well.

I value Run D and Agility very highly. And, depending on some things with my defense, I value M2M/Play Diag or PRT/PRS next. In this case, I was looking for a good cover ILB.

It's good to learn something new. Frankly, I usually only use DA as a way to find guys quickly that I like, then I'll sort through them myself and make some choices based on my thoughts about the position and the offense/defense I'm trying to go with. I'm still baffled why I went this route for this one pick :)

[Nemesis]

It seems to value run D a LOT.

It's because the Run Defense bar has the biggest influence on overall rating.

[Guest]

Fwiw, Zone seems to be more useful/effective than man or bnr in LBs.

[Aston]

I haven't followed this discussion on slotting but fwiw, my understanding has been that the D?, VG, G, etc numbers were solely based on combines and bars (from draft_personal).

[Guest]

I thought the same thing, Rando. He wasn't that good, IMO. But, I guess the weights I use affect the slotting as well.

Yeah, it could be the weights. Or the coaches. Or probably a bit of both. We'd need multiple people's files and multiple people's DA weights to figure it out, and I for one really don't want to know.

I like the uncertainty of the new draft. Even with all this data and all these tools, every time my pick comes up I'm left wondering and second guessing myself.

[Guest]

I haven't followed this discussion on slotting but fwiw, my understanding has been that the D?, VG, G, etc numbers were solely based on combines and bars (from draft_personal).

Looks like it. Editing the weights change the Comb. score as well as Rate score and their overall position in the "slot for draft", but D? and whatnot numbers always seem to stay the same.

I'm guessing they're something like Ben's combine analyzer thingy on steroids. It just looks at the raw numbers and outputs its best guess on where that player will end up as an overall based on previous players or something. While everyone should see the same combines, the bars will be different and so then will those scores.

The weights and the chemistry stuff then let you narrow down and give preference to specific skills, positions or zodiac signs you are looking for. What they give you might not be the best players overall, but they are the best according to what the DA thinks you are looking for.

For example, if I cranked up the "run blocking" bar weight for all the lineman positions, it would then preferentially sort those linemen who, according to my coaching staff, had high run blocking bars in the draft_personnel file. The problem is, if my coaching staff isn't very good at analyzing offensive linemen, the results DA aren't going to match what I wanted after the players are revealed for what they really are.

I might be better off cranking up the "40 time" combine weight or a combination of both weights just as a sanity check, but even then there are a wide range of "middling" 40 times that can correspond with a wide range of run blocking scores.

So you need to look for something else. Some other tells. And if all else fails, turn the targeting computer off and just use the force.

In the end it can be a usefull tool, but you have to remember that not every problem is a nail.

[Aston]

Circling back a bit, the width of the orange bars might be completely determined by the width of the blue bars. Just a guess.

[Guest]

And if all else fails, turn the targeting computer off and just use the force.

[Delandis]

My best draft so far has been when I was fresh to the game. I drafted without correlations and masking pairs and every other method to crack the code. I think the game is designed to be evaluated at face value. If he has good bars and combine you draft him.

I need to get back to that.


Sent from my iPhone using