NOTICE: This league is using the BLEEDING EDGE game engine. For more information, click here.

The new user interface is in preview!

Want to check it out? Click here! (If you don't like it, you can still switch back)

League Forums

Main - Beta Chat

Re: 2035 Draft Class

By WarEagle
12/14/2017 2:20 pm
jdavidbakr wrote:
WarEagle wrote:
jdavidbakr wrote:

Technically, we won't know if it works for at least a few seasons.

At which point these players would have probably already been cut. How is this any different than what currently happens?

jdavidbakr wrote:

And probably won't really know if it works until the code lands in some non-beta leagues where we don't have quite the experienced drafters as setherick mentioned.

So the goal is for this to only work in leagues where the owners don't know what they are doing?


I'm not sure what you are wanting it to be - either we have it where there is a meaningful amount of ability to predict how good a player is, which makes higher draft picks more valuable, or your draft position is meaningless and the ultimate value of the players is completely random. The goal of the algorithm is to increase the probability that a late round draft pick will be a gem - it doesn't even mean that every draft will have such a gem, and definitely doesn't mean that every team will get such a gem, it just means that those gems have a better chance to exist than previously.


I guess I just misunderstood the whole purpose behind any changes you were making. I haven't been paying much attention lately.

Of course I do not want all draft picks to have the same chance of getting a good player.

I thought what you were doing was specifically (and ONLY) related to players picked in the late rounds (5-7), and that your changes would result in a few (not a lot or none) players picked in those rounds being revealed as good>great players after TC (including improvements to the non-improvable attributes such as speed/strength/int).
IE: "We didn't know what we had until we got him into camp and saw ...".

I thought the main purpose was to make the later rounds more interesting and worth keeping/making those picks, but that hasn't even been true for the first round lately.


Re: 2035 Draft Class

By jdavidbakr - Site Admin
12/14/2017 4:20 pm
WarEagle wrote:
jdavidbakr wrote:
WarEagle wrote:
jdavidbakr wrote:

Technically, we won't know if it works for at least a few seasons.

At which point these players would have probably already been cut. How is this any different than what currently happens?

jdavidbakr wrote:

And probably won't really know if it works until the code lands in some non-beta leagues where we don't have quite the experienced drafters as setherick mentioned.

So the goal is for this to only work in leagues where the owners don't know what they are doing?


I'm not sure what you are wanting it to be - either we have it where there is a meaningful amount of ability to predict how good a player is, which makes higher draft picks more valuable, or your draft position is meaningless and the ultimate value of the players is completely random. The goal of the algorithm is to increase the probability that a late round draft pick will be a gem - it doesn't even mean that every draft will have such a gem, and definitely doesn't mean that every team will get such a gem, it just means that those gems have a better chance to exist than previously.


I guess I just misunderstood the whole purpose behind any changes you were making. I haven't been paying much attention lately.

Of course I do not want all draft picks to have the same chance of getting a good player.

I thought what you were doing was specifically (and ONLY) related to players picked in the late rounds (5-7), and that your changes would result in a few (not a lot or none) players picked in those rounds being revealed as good>great players after TC (including improvements to the non-improvable attributes such as speed/strength/int).
IE: "We didn't know what we had until we got him into camp and saw ...".

I thought the main purpose was to make the later rounds more interesting and worth keeping/making those picks, but that hasn't even been true for the first round lately.




I did give a random volatility boost to the later rounds, I don't know how many players that impacted. I'm guessing it will not be a whole lot, I'm interested to see if any happened. It's possible that the frequency is too low - I'd rather have it too low and gradually turn it up than have it too high and flood the talent pool too much.

Re: 2035 Draft Class

By setherick
12/14/2017 4:21 pm
Most of the ones I saw that were impacted by this "boost" busted hard. I'll try to dig up some examples later.

Re: 2035 Draft Class

By WarEagle
12/14/2017 8:11 pm
I don't think a simple volatility boost is going to turn any of these late round scrubs into "gems", even if it was given to all 96 players.

Re: 2035 Draft Class

By setherick
12/14/2017 8:47 pm
WarEagle wrote:
I don't think a simple volatility boost is going to turn any of these late round scrubs into "gems", even if it was given to all 96 players.


This is absolutely true because so many key attributes don't change. So all of us have been trained to draft the fastest or strongest at a position, and then hope that high vol makes the player better. Until Speed, Strength, or Intelligence improves, there will be no late Pro Bowlers or HOF'ers because they will have been taken in rounds 1-4. This was one of the reasons that JDB tried the obfuscation, but that was also kind of lame because it didn't make logical sense. The only way I could see this working out is if players had listed ratings and ratings only available to the person's owner so like Speed versus Game Speed, etc.

Re: 2035 Draft Class

By WarEagle
12/14/2017 9:06 pm
There should be late round picks who are randomly selected to turn into "gems", and these players should have all attributes boosted significantly in TC (not just have volatility boosted).

Their ratings for Int / Speed / Strength / Courage / etc. can go from a potential of 45 to 95 (for example) in the first TC. Otherwise, there will never be any and this whole idea might as well be scrapped.

Re: 2035 Draft Class

By mardn72
12/14/2017 10:07 pm
I actually think the changes in this draft to static attributes made a difference, but had some unintended consequences too. Since more of the later round players creation value was spent on the static attributes, the dynamic ones were generally less. Especially problematic is the smaller gap between current and potential attributes. Even with high volatility, a small gap allows the rating to move only a little. As JDB has explained, the closer the actual is to the potential, the less the potential will move.

I found plenty of players with enough good (not great) static attributes, but with little available growth range. For example, a player that has 15/25 in M2M, even with 100 vol could top out at 35 or 40. I’d like to see volatility be able to push beyond these limits a bit more.

I also think we should figure out what we really expect from these later rounds. When I did research in the past, the last few round really only produce about 1 starter per round each season, and only a handful of hall of gamers in the history of the NFL. So I agree with JDB that we need to carefully tweak it to keep good-to-great player rare in the late rounds. (They were basically non-existent before).

Re: 2035 Draft Class

By setherick
12/15/2017 12:13 am
Beta 87 had quite a few late round boomers that will be starters or solid backups for their teams.

5.32 - https://beta87.myfootballnow.com/player/3877
6.1 - https://beta87.myfootballnow.com/player/3918 - Not ideal speed or strength for a TE. I would have moved this guy to WR.
6.13 - https://beta87.myfootballnow.com/player/4021 - TE I converted to OL.
7.20 - https://beta87.myfootballnow.com/player/3951 - This could could be a WR 3 on many teams.
7.31 - https://beta87.myfootballnow.com/player/3831 - Valuable in a RB rotation. I would have loved to get this guy at 7.31. But my 7th round pick went -20.

Also, by my weights, 11 of the top 96 players in this draft not counting kickers and punters came from rounds 5, 6, and 7. The fact that two of these players ended up on my team probably says more about my weights than my drafting.

5.4 - https://mfn1.myfootballnow.com/player/13845
5.15 - https://mfn1.myfootballnow.com/player/13801
5.21 - https://mfn1.myfootballnow.com/player/13613
5.22 - https://mfn1.myfootballnow.com/player/13885
5.25 - https://mfn1.myfootballnow.com/player/13843
5.28 - https://mfn1.myfootballnow.com/player/13892
5.31 - https://mfn1.myfootballnow.com/player/13891
6.13 - https://mfn1.myfootballnow.com/player/13805
6.15 - https://mfn1.myfootballnow.com/player/13810
6.31 - https://mfn1.myfootballnow.com/player/13827
7.27 - https://mfn1.myfootballnow.com/player/13852
Last edited at 12/15/2017 12:24 am