If You Build A Better Mousetrap Or Scoring System For The Den Draft Prediciton Contest

The Den
1

…, the world will beat a path to your door.

Every year The Den cobbles together a Lions draft prediction contest. Every year the scoring system du jour confuses people. Hell, du jour confuses people. Every year it doesn’t matter for the vast majority of Densters because Brad’s prime directive is to F*** with our contest and, only secondarily, to draft good players.

So do you have any simple ideas on how to score the Lions draft prediction contest? This could be helpful for whoever grabs the bull by the horns at the last second.

2

We’ve tried it so many different ways and the problem is that it’s easy to manipulate a simple points system. So it gets complicated to keep people from manipulating the system. M

People hedge (listing multiple similar players, clustering positions, draft rounds etc.). The trick is to reward precision without making a scoring system that is easily manipulated

You’d think you could put together a simple system but you can’t without someone trying to manipulate it or finding the loophole.

I asked AI to design a simple scoring system that is impossible to game and hedge the system.

AI came up with a scoring system very similar to the complicated one people struggle to understand.

1 Like
3

I’d say:

The Lions pick, multiplied by the position he plays, divided by the round, added to your picks position, squared by the round it is. Then take your next picks position, divided by the round he went, subtracting if the team who actually took him is in our division, halved by Mel Kipers final mock position of the player. Then add all your numbers up, but subtract by what round and team he was taken by.

Easy.

3 Likes
4

KISS:
Name ten (or five, eight, whatever) players you think the Lions will draft.
Earn a point for each correct player.
Forget about rounds, forget about positions, forget about how many picks we have.
Just one point for each correct player.
There will be no debates on how to score this.

5 Likes
5

Judging from past predictions and actual draft, I’d say that it matters not. As a group…we suck. :face_with_crossed_out_eyes: :sweat_smile:

6

Losing Oh No GIF by Star Now

7

True, the scoring would be easy.

But to @Air2theThrown 's point, “the problem is that it’s easy to manipulate a simple points system.” Like use 5 of your 10 players likely to go at #17, and maybe 3 or 4 likely to go at #50, because later rounds are so hard to get right.

Maybe a system with one pick at #17, 2 picks at #50, 3 at #118, etc? But that’s leaves KISS territory…

1 Like
8

Here’s what AI suggested for me. A little convoluted, yes, but not overly complicated.

Screenshot 2026-04-14 at 11.39.46 AM
Screenshot 2026-04-14 at 11.39.46 AM1418×674 27 KB

9

I would expect alot of people to do that, because it is easier to guess the early rounds. but then they are giving up any chance at more than one or two points. But so what? If you can totally lose the concept of “which round”, its really not gaming the system.
Whereas the real draftniks might be braver to pick some later round people that are very Lionesque.

That may have worked in the past but Brad trades picks like he’s on Wall St. Any concept of “what pick number (e.g. 50)” is likely to get complicated. Maybe pick the player, and then, separately, pick what position he picks in each round?
e.g. Spencer Fano, Akeem Mesidor, Jacas, Rodriguez, etc. (first just the player), any correct name gets a point,
then, separately,
e.g. 1st round (or bettter yet, first pick) OT, second pick Edge, third TE, fourth LB, etc… one point for a correct position/pick placement. You’d still get credit for picking a LB with the fifth pick, if that’s what you predict.
Getting a little more complicated but I think its not that hard.

10

I like the KISS suggestion. Takes the brad trading out of concern.

11

The problem is that often the winner only has one correct pick. With scores so low, that one or two points could easily be the difference.

Also I just think it’s kind of a bunk contest if people just list 10 1st rounders. That’s not a real draft prediction contest, that’s a “guess the 1st rounder and get 8 cracks at it” contest.

12

If you wanna do the 10 player model… Use a composite big board… Then allow two top 50 ranked players, four players ranked 50-150, and four players ranked 150-250

13

That would work. Or you could give a penalty if the pick wasn’t made close-ish to the guess. ±60 spots for day 3 guys, or something like that.

14

that happened last year. The pick was Teslaa. Other than that, its usually more than one.

In 2024, someone had rakestraw, vaki, and mahogony.

In 2023, it was two - some had branch and campbell, some had gibbs and campbell, some had Hooker and Branch.

in 2022, it was also two. Hutch and rodrigo, and Hutch and Kerby.

More often than not, the later picks are making the person win. So it wouldn’t be good strategy to pick 10 first rounders.

15

Fair enough, I stand corrected. (Also I know I won 2021 with Sewell and Alim. So another with 2). But if you got the 1st and 2nd rounder right, it would still be a pretty good strategy, especially since a lot of those guys could theoretically be either, like Ponds, Parker or Lawrence.

You’re probably right that it would work out fine in the end, it just doesn’t feel like as interesting of a contest to me. Though I am admittedly a deep diver. I still suspect the vast majority of high scorers will just put a bunch of players we talk about all the time.

16

I hear you and I think deep diving ought to somehow be rewarded. (lol).

I dunno, maybe have the top 32 or so players from the Beast listed and say you can only at most pick three of these. It would be easier to verify if the players were listed up front, rather than depending on the honor system to verify them.

1 Like
17

BTW, @Air2theThrown , are you gonna herd cats and spearhead the contest this year? I know you often do it.

18

It’s just for fun anyway. Keep a similar as always point system, and have a points winner. Then have a vote for the people’s choice. I know I’ll always vote for a guy who hits on a later pick vs hitting on the first pick. With the caveat, they didn’t trade down 14 times and had 22 picks :rofl:

Or just not really care and have a thread with what you think will happen. And just discuss it.

19

Notsofastso!
Looks like you’re it!

20

@Air2theThrown is The Man and it’s his thang until he says it isn’t.

Here’s my better mousetrap (with or without the alternate pick concept) which is KISS adjacent:

You score 2 points for the right player at the right slot or slot range due to small trades.

You score 1 point for the right alternate player at the right slot or slot range.

At #17, there are no alternate players and the slot range is #16 to 18.

At #50, choose 1 alternate player and the slot range is #47 to 53.

At #118, choose 1 alternate player and the slot range is #114 to 122.

At #128, choose 1 alternate player and the slot range is #124 to 132.

At #157, choose 1 alternate player and the slot range is #152 to 162.

At #181, choose 1 alternate player and the slot range is #176 to 186.

At #213, choose 1 alternate player and the slot range is #208 to 218.

At 222, choose 1 alternate player and the slot range is #217 to 227.

Any pick outside the slot range parameters, gets zilch, zero, nada. Brad is messing with The Den again.

Examples: At #50, you pick player A and have player B as an alternate. If Brad trades back to #53 and selects player B, you get one point, the same as if Brad hadn’t traded at all.

If Brad trades down to #47 and selects player A, you get 2 points, the same as if Brad hadn’t traded at all.

If Brad trades up to #46 or down to #54 and picks player A or B, you get zilch, zero, nada…and a reason to b*tch.