Pick Em Point Totals

[Jamar Howard 4 President]

I assumed you would put 4 points on the highest percentage and 3 on the next highest. Wouldn't this maximize the expected value? In this week though it would have meant Ken only scores 6. With retroactive "predictions" I think you need to spell out the methodology.
The first 3 weeks how did you distribute the points?

Sorry for the nitpicking, but I'm interested in seeing how Ken does and I want to make sure it is fair/accurate.

Not nitpicking. Good question. That is a fair point that interpreting %'s differently could lead to assigning different point values. As you can see, I've been taking a more (spread your risk) approach. Your proposal to give 4 points to the highest %, regardless of how little difference between it and the next highest, is interesting. I believe you are correct that mathematically, it would be *slightly* better to do it that way.

Here's the last 6 weeks, calculated as I've choosen to post them (spread your risk), and then recalculated via the 4,3,1,1,1 system every week.

Week 1 - 6 ... 8
Week 2 - 10 ... 10
Week 3 - 8 ... 7
Week 4 - 5 ... 5
Week 5 - 9 ... 9
Week 6 - 7 ... 6
Total ... - 45 .. 45

To-date, both methods have earned the same 45 pts. I can calculate both ways going forward. We will call it Ken1 and Ken2. I have also been posting the %'s each week, so if anyone comes up with a Ken3, we can tally that as well.

[ShockerFever]

Wow, can I get 13 picks for each game too?

[Shocker1976]

Here are Ken's Week 6 projections and results

Projections:
ISUb - 2 (69%)
EU - 2 (70%) - WRONG
WSU - 1 (51%) - WRONG
LOY - 2 (67%)
UNI - 3 (74%)

Ken earned 7 pts in week 6. Now has 45 total and is tied for 3rd overall.

JH4P I do not want to include these predictions as they reflect "an after the fact" assignment of points per game. I want you to participate in this exercise using KenPom data with you assigning points per game prior to these games being played. I recognize including the first four picks is an exception to this mindset but I prefer to make such an exception a one time event.

[moshock]

JH4P I do not want to include these predictions as they reflect "an after the fact" assignment of points per game. I want you to participate in this exercise using KenPom data with you assigning points per game prior to these games being played. I recognize including the first four picks is an exception to this mindset but I prefer to make such an exception a one time event.

I believe that he is assigning points based on % chance of winning

[Jamar Howard 4 President]

JH4P I do not want to include these predictions as they reflect "an after the fact" assignment of points per game. I want you to participate in this exercise using KenPom data with you assigning points per game prior to these games being played. I recognize including the first four picks is an exception to this mindset but I prefer to make such an exception a one time event.

Feel free to dump my "spread the risk" method as yes, I didn't fully document it and folks could theoretically disagree how to apply it using Ken's data. However, this is why I started calculating using 4,3,1,1,1. There is zero opinion. It is a pure regurgitation of Ken's data using a constant method. Give his highest % game a 4, the next a 3, and then all 1s.

Why not include this method, especially because it is just for comparison purposes? Obviously Ken isn't really playing, so he doesn't qualify to "win" this contest, but it is an interesting point of reference.

[Shocker1976]

I believe that he is assigning points based on % chance of winning

I know this but I thought having him make the assignment prior to tip off per our rules is important. However, I will amend the spreadsheet to include all of his picks (another poster although a non-participant has also indicated support for such inclusion) should another contest participant make such a request.

[Shocker1976]

Feel free to dump my "spread the risk" method as yes, I didn't fully document it and folks could theoretically disagree how to apply it using Ken's data. However, this is why I started calculating using 4,3,1,1,1. There is zero opinion. It is a pure regurgitation of Ken's data using a constant method. Give his highest % game a 4, the next a 3, and then all 1s.

Why not include this method, especially because it is just for comparison purposes? Obviously Ken isn't really playing, so he doesn't qualify to "win" this contest, but it is an interesting point of reference.

You make a good argument, so I will amend the spreadsheet.

[Shocker1976]

Amended to include KenPom data.

Week	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18	Total
Top Scores
capnkirk	5	10	7	7	10	10													49
valleyclimber	9	10	7	5	9	9													49
JH4P/KenPom	6	10	8	5	9	7													45
Play Angry	8	9	7	6	10	5													45
shocks771	6	10	7	6	8	8													45
im4wsu	8	9	7	5	9	6													44
EliShocker	5	9	7	6	10	6													43
BAShocker	4	8	8	6	10	6													42
Onegreatracer	6	7	9	7	9	4													42
shocks02	5	9	6	4	9	9													42
WSU and Reds Fan	5	8	6	7	9	7													42
wufan	7	9	6	5	7	8													42
Aargh	7	9	6	3	10	6													41
AndShock	6	9	6	5	8	3													37
bleednblackngold	4	8	6	7	8	5													38
btp1029	5	8	6	5	6	7													37
B1ShockFan	5	8	6	5	8	8													40
Calshockfan	5	9	6	5	8	5													38
ccrunner101	5	10	8	5	8	5													41
COS Shocker	5	9	5	5	7	8													39
dregn	7	8	7	5	8	6													41
DUShock	2	8	5	4	8	4													31
dwbarcl	5	9	4	7	9	3													37
eeshocker	4	10	5	6	9	6													40
goob65	5	9	6	6	7	5													38
Horn28Clem30	4	6	6	6	9	8													39
molly jabali	4		7	3	9	9													32
moshock	5	10	5	6	8	5													39
MoValley John	4	9	5	6	8	5													37
newshock1234	5	9	7	5	8	5													39
pie n eye	5	9	7		8														29
pogo	5	9	6	4															24
proshox	3	9	8	8	9	3													40
Ricardo del Rio	4	8	5	4	7	4													32
rrshock	5	9	5	5	9	7													40
SB Shock	6	9	5	5	9	6													40
Shockadelica	6	7	6	4	8	8													39
ShockdaWorld	5	9	5	4	9	5													37
ShockerDave	4	7	6	5	7	5													34
ShockerDropOut	5	7	8	5	10	4													39
shockerfanmas	4	9	5	5	8														31
ShockerFever	5	8	7	5	8	8													41
Shocker1976	6	10	6	4	10	5													41
Shockerfan34	6	8		4	8	7													33
ShockerMas	4	8	7	2	10														31
shockerofandover0943	4	10	8	5	9	4													40
ShockerPhi	5	7	5	4	8	4													33
shocker22	6	8	7	7	6	5													39
shock-it-to-me	6	7	5	5	8	7													38
ShockTalk	4	7	5	5	9	8													38
shocktheheart	6	9	6	3	9	6													39
shox_engineer24	3	10	6	5	10	7													41
ShoxFan7	5	9	4	4	8	5													35
SHOXJOCK	4	9	6	5	8	6													38
SHOXMVC	4	6	6	4	4	8													32
shoxxfmr	4	10	8	4	8	7													41
SumShocker	5	8	7	4	8	6													38
Triticum aestivum l	5	9	7	4	8	8													41
vancedave56	3	7	6	6		5													27
WheatShock	6	8	6	5	9	5													39
WuShock16	3	9	6	4	6	7													35
wushock27	6	8	4	8	10	4													40
Zardoz	5	9	6	6	7	4													37
1972Shocker	4	6	6	7	9	4													36

[im4wsu]

Here are Ken's Week 6 projections and results

Projections:
ISUb - 2 (69%)
EU - 2 (70%) - WRONG
WSU - 1 (51%) - WRONG
LOY - 2 (67%)
UNI - 3 (74%)

Ken earned 7 pts in week 6. Now has 45 total and is tied for 3rd overall.

How does KenPom assign the points? Why wouldn't they all be two OR the logical alternative, UNI 4, EU 3, and the remainder one each, according to the percentage odds? What is shown appears to have arbitrarily assigned points.

[ShockerFever]

It's kinda funny when Jamar avoids the Fever comparison spreadsheets when he loses the round. How funny. Glad he can take it on the chin like a man.

[Jamar Howard 4 President]

How does KenPom assign the points? Why wouldn't they all be two OR the logical alternative, UNI 4, EU 3, and the remainder one each, according to the percentage odds? What is shown appears to have arbitrarily assigned points.

He doesn't. He just provides "team X has a Y% chance of winning".

I never intended to play this game. My name was brought up and I was asked to post Ken's thoughts. I'm providing this info as a favor, and yes, having a little fun with "he-who-shall-not-be-named" at the same time. Today, it was (correctly) brought up that there could be differing opinions on how to assign points this way. After additional thought, I realized that not only is 4,3,1,1,1 mathematically superior, it also removes opportunity for disagreement on how to assign points. 4 to the game with the highest %, 3 to the next, and 1 to each of the rest.

See post #61 of this thread for more of my thoughts.

[Ricardo del Rio]

Someone has hacked the ShockerNet site and changed my picks.

Investigation, please.

[MoValley John]

Someone has hacked the ShockerNet site and changed my picks.

Investigation, please.

Mine, too. I didn't pick that garbage!

[im4wsu]

He doesn't. He just provides "team X has a Y% chance of winning".

I never intended to play this game. My name was brought up and I was asked to post Ken's thoughts. I'm providing this info as a favor, and yes, having a little fun with "he-who-shall-not-be-named" at the same time. Today, it was (correctly) brought up that there could be differing opinions on how to assign points this way. After additional thought, I realized that not only is 4,3,1,1,1 mathematically superior, it also removes opportunity for disagreement on how to assign points. 4 to the game with the highest %, 3 to the next, and 1 to each of the rest.

See post #61 of this thread for more of my thoughts.

Sorry, I must have either read the #61 and related posts while I was asleep or somehow skipped over them going to first unread. My bad.

[MadaboutWu]

He doesn't. He just provides "team X has a Y% chance of winning".

After additional thought, I realized that not only is 4,3,1,1,1 mathematically superior, it also removes opportunity for disagreement on how to assign points. 4 to the game with the highest %, 3 to the next, and 1 to each of the rest.

Yes, hedging you are making subjective calls and so doing that after the fact it is hard to be unbiased even if you are trying to be.

Going with 4,3,1,1,1 maximizes points though increases variance. The expected value difference is of course rather slight, in round 5 it was .01 and the actual picks it produced 6 points instead of the 7 which your particular hedge scored.