CSNbbs

Full Version: Tweak Pick'Em?
You're currently viewing a stripped down version of our content. View the full version with proper formatting.
On the Pioneer Football board, we're predicting final scores. Results are based on w-l record first, then victory margin pts off (predicted margin minus actual margin if correct winner picked, predicted margin plus actual margin if wrong winner picked...lower number the better), then score accuracy points (1pt for correct predicted score, 1/2 pt if within 3...higher number the better).

http://pflfan.proboards26.com/index.cgi?...1188519549

Do we want to do something similar for A-Sun Pick'Em? Last year's confidence level points worked fine, but this scoring stuff is kinda interesting. Victory margin pts off will work fine for hoops. Score accuracy pts work well for football, but I'm not as sold on it for hoops. However, we could do 1 pt for within 2 and 1/2 pt for within 5. It'll be more complicated to show results, but it might be interesting and will be another measure of pickers' accuracy.

Feedback please!!
I think I like the idea of confidence level better like we did last year. That's just me though, what does everyone else think.
Instead of predicting final scores, we can predict margins of victory. If you think JU will beat Lipscomb by 5, post your pick as....

LIP @ JU 5

This can be interesting in those possibly lop-sided non-conference games, whether its Stetson at Florida or Clark Atlanta at Kennesaw.

Just a thought. Either way will be fine. I think we can nix predicting final scores, though.
KSU at ETSU 35

:)
I like that idea OC.
Nick M Wrote:KSU at ETSU 35

:)

Funny...I wondered if a KSU fan or an ETSU fan would do this first. Note that I purposely picked Lipscomb and JU so I wouldn't offend any posters!

So on NickM's example...if Kennesaw actually won by 33 or by 37 he would be 1-0 / 2.

If KSU won by 1, he'd be 1-0 / 34.

If ETSU won by 12, he'd be 0-1 / 47.

Like strokes over par in golf, the lower number...the better.

What do y'all think?

Also, do we do the non-conference games and then start over from scratch once conference games begin? (I say "yes".)
OrangeCamel Wrote:Funny...I wondered if a KSU fan or an ETSU fan would do this first.

Right on time then, clockwork baby!
I like the score idea. Makes things interesting.
I like this new idea. I think we should do it! I am down with the non-conference games as well. Keep it interesting early on before a-sun action.
OOC Pick'Em is winding-down and the "score margin pts off" are getting astronomical. They show how bad we are and, frankly, it's all a blur and is kinda boring. (Of course, the W-L records are the main basis of the results.)

Instead of "score margin pts off" for the upcoming conference games, what if we do "score accuracy pts"? We'll still use the same formula (predicted margin less actual margin if correct winner picked, predicted margin plus actual if wrong winner picked). "Score accuracy pts" will be rewarded for good picks, as follows: (more pts, the better)

3 pts for a perfect prediction
2 pts for 1 pt off using the formula
1 pt for 2 or 3 pts off using the formula

Here's a comparison, using OOC Round #1 with those early upsets:

"score margin pts off"

MercerFan 9-3 / 96
BucFyre 8-4 / 119
ETSUfan1 8-4 / 127
kr983 8-4 / 137
OrangeCamel 8-4 / 139
Buccaneerlover 7-5 / 144
KennesawBasketball 7-5 / 157
Collinr01 7-5 / 159

vs "score accuracy pts"

MercerFan 9-3 / 8
ETSUfan1 8-4 / 5
OrangeCamel 8-4 / 4
BucFyre 8-4 / 3
kr983 8-4 / 3
Collinr01 7-5 / 5
KennesawBasketball 7-5 / 3
Buccaneerlover 7-5 / 2

I think the smaller numbers with a 6 pt spread (vs. the larger numbers with a 63 pt spread) are more interesting and more competitive. If you're 4 pts off or 50 pts off, you'll get "0" score accuracy pts either way. Technically, it might not measure accuracy as well, but it provides a twist that might be more fun. (Also, it will be a heck of a lot easier on me as it will eliminate a lot of math.)

What do you think?
I liked the confidence level thing, but am up for whatever...

FWIW, KSU @ ETSU 40 with a 1 confidence level. 110-70
That's pretty good you're giving us 70... normally when a college team scores +100 it's not that close. In reality it'll be a 5 point game in KSU's favor but whatever. That's what the other board is for.
Orange Camel --- I like the new system you listed there a lot! That would really make me think about the final score more, trying to get it perfect. Instead of just ballparking it. Anything to make it easier for you too, as we all enjoy it a lot.

Thanks
Check this out.....

OOC cumulative "all games" thru Round #12

OrangeCamel 94-28 / 1213 46
MercerFan 94-28 / 1302 43
ETSUfan1 89-33 / 1372 38

"score margin pts off" has a range of 159 pts (1372 minus 1213) between the three pickers.

"accuracy pts" (in red) has a range of only 8 pts (46 minus 38), w/ only 3 pts between OC and MF....much more interesting, I think!
Final Round thru 12-28 (proposed "accuracy pts" in red)

andone 2-0 / 3 4
ETSUfan1 2-0 / 3 3
MercerFan 2-0 / 6 2
OrangeCamel 2-0 / 7 1
kenpom.com 2-0 / 17 1
Buccaneerlover 1-1 / 20 0

A couple of issues with the accuracy pts are exposed here....

Both, andone and ETSUfan1 are 3 "score margin pts off", but andone gets one more "accuracy pt" since he picked the Campbell game perfectly. Is that fair?

Also, although I'm 10 "score margin pts off" better than kenpom.com, we're tied with 1 "accuracy pt". kenpom's way-off Kennesaw prediction vanishes into thin air.

Which way do you all prefer? We can do either. The purist in me says stick with the "score margin pts off", but if the "accuracy pt" way will make it more interesting and fun, we can to that. Let's decide!
Reference URL's