New Weighting Proposal For Yearlong Contest

OK...we have had some time to see how the numbers work out.
It's time to get official here and vote yes/no on a proposal.

To that end, I've crunched the numbers a lot, and think I have what could be a fair proposal...that gives everyone a chance and does not really allow for a runaway.

To be fair, we need to have several factors involved in the weighting.

I believe the factors that should be considered are as follows:

1. Time on TrackDollar - the longer you have been here, the more time your bills have had to circulate.
The average life of a bill is said to be 18 months...so we only consider newbies, yearlings, and veterans.

2. - Amount of bills actually Entered - Obviously, the more bills you have out there, the better your chances.

3. - Track Points - Obviously, those with higher Track Points either have more entries, more hits (probably more time in, too) or a combination of any/all of the above. Thus, they have better chances.

4. Hit Rate - Obviously, the higher your Hit Rate, the better your chances are.

5. Location - This is a special thing. I want to give Dave a little boost here, as he has a disadvantage over the rest of us. He lives in an area where there are HUGE counties. You could put four of our counties into any of the counties surrounding Dave. anyone living out West, where there are HUGE Counties...would get a boost factor of 1.4
This makes sense, because, currently, I am in PA - 67 Counties.
Markus and Tim are in VA - 134 counties (134/2 = 67) and they are in different areas, Markus in Southern VA, Tim in Northern VA...so that balances. Markus has a bit of an advantage in TX, but that does not appear to have made much impact.
Dave, on the other hand, has but 16 counties in his entire state...and is surrounded by NM and CA, and UT...all of which also have HUGE counties. He could get five hits all in the same county...that, for us, would yield 4-5 counties! so I propose this boost for dave...and anyone else who joins and lives out West where counties are huge.

SO...let's get some actual numbers attached to these factors!!

Time on TD -
If joined in 2011 - factor 1.2
If joined in 2010 - factor 1.0
If joined prior to 2010 - factor 0.9

Entry -
1-2500 - factor 1.10
2501-5000 - factor 1.08
5001-7500 - factor 1.05
7501-10000 - factor 1.03
10001-15000 - factor 1.0
15001-20000 - factor 0.98
Over 20000 - factor 0.95

TrackPoints -
1-2500 - factor 1.10
2501-5000 - factor 1.08
5001-7500 - factor 1.05
7501-10000 - factor 1.03
10001-15000 - factor 1.00
15001-20000 - factor 0.98
Over 20,000 - factor 0.95

Hit Rate -
1.x% - factor 1.15
2.x% - factor 1.10
3.x% - factor 1.08
4.x% - factor 1.05
5.x% - factor 1.03
6.x% - factor 1.00
7.x% - factor 0.98
8.x% - factor 0.95
9.x% - factor 0.93
10% and up - factor 0.90

West Boost -
West - factor 1.4
East - factor 1.0

Given these numbers...these would be the weightings of each player currently in the contest...

(coming up gotta do the math!!)

This would yield the following results:

Kalisiin weighting = 1.143072
dctim weighting = 0.944784
Markus weighting = 0.8470728
Dave weighting = 1.323

To make the math easier, these would all be rounded to 2 decimal places.

Kalisiin = 1.14
dctim = 0.94
Markus = 0.85
Dave = 1.32

That would make the following scores currently...

Kalisiin = 43.32 (38 hits)
dctim = 16.92 (18 hits) - if he had 38 it would be 35.72
Markus = 11.9 (14 hits) - if he had 38 it would be 32.3
Dave = 9.24 (7 hits) - if he had 38 it would be 50.16

Can we have a vote or comments or alternate suggestions?
We really need to have a final proposal everyone agrees with, so that maintaining the standings will become much easier for me.

Anyone want to say yes/no to this idea...or propose something else?

Incidentally, this contest is looking, right now, like a bit of a runaway with me in the lead...but, since I have over double what anyone else has...this sorta makes sense. But it DOES make me just a bit uncomfortable...would you guys hurry up and get some more freaking hits??

I hope to get a chance to crunch the numbers myself in the next day or two - just for fun.

With that said, I think Markus has a few more counties that he has not yet reported, so those should be a help to him.

I have gone on a small new-county-streak over my last 7-10 hits, I've gotten several new counties. However, both my hits and my entering have reduced over the last few months. I will be spending a few days in March in the western part of the county just to the west of me - Perhaps I will be able to score a few new VA counties that way.

I am not too concerned that you have the lead that you have. You have picked off the "easy" counties, new counties later in the year may be tougher for you.

dctim wrote:

I am not too concerned that you have the lead that you have. You have picked off the "easy" counties, new counties later in the year may be tougher for you.

Exactly.

But it looks a little bad right now, and makes me a bit uncomfortable.
I don't want anyone thinking I am rigging this thing to my advantage.

Because I want this to be a yearly event.

But we have to come up with a weighting formula that works. I think what I proposed above is pretty good. It takes many factors into account, and gives a boost to newbs...and a bit of a hindrance on long-termers, so that they don't score a runaway.

Works for me if it works for you....

So we have Dave and myself. Need to hear from Tim and Markus.

Here is my table plus analysis:

Name	Hit	Time	Entry	T pts	Hit %	West?	Avg	Score
Kalisiin	48	1.00	1.08	1.08	0.98	1.00	1.028	49.34
Markus	46	0.90	0.98	0.98	0.98	1.00	0.968	44.53
dctim	22	0.90	1.08	1.08	0.90	1.00	0.992	21.82
Diamond Dave	8	0.90	1.00	1.00	1.05	1.40	1.070	8.56

I did a simple average of the above ratios for the "Avg" column.

Kalisiin and I will likely break the 5,000 Bills Entered threshold during the year. Kalisiin will likely break the 8 % Hit rate threshold. Dave will likely break the 15,000 bills entered threshold, if not the 20 K with Markus.

With that said, for us all to be statistically tied ...

If Kalisiin's had 50 Hits, Marcus would need 53, I would need 52 and Dave would need 48 given the above ratios.

I am pretty comfortable with this. It shows that the ratios are about as fair as we can make them right now.

My opinion right now is that if Marcus reports his hits, the contest is his to loose.

Allright then. We have three of four players...a majority agreeing to the new weighting...so, unless Markus has some serious objection, we will go forward with that plan.
We will give Markus two days from now to voice any serious objection - if he has any. If no hear, then we go forward.

As I said, I tried my best to crunch numbers that would give a slight boost to newbs...and a slight hindrance to old hands...so that we would not have a runaway...nor would any one person be totally out of the running.

I think this works, given the figures Tim has provided above...so, like I said, unless Markus has any major objection, then we go forward.

Now the next question becomes...do we want first place to win everything? Should there be two equal winners? Should there be a larger first prize and a smaller second-place prize?

For this year, I say it is just first prize and that is it...but maybe next year...following years...if we get more participants and a larger pool of players...this is something to consider.

With only 4 people in the contest, I say winner takes all. Should we get a deluge of new participants, we can re-evaluate.

dctim wrote:With only 4 people in the contest, I say winner takes all. Should we get a deluge of new participants, we can re-evaluate.

That, too, was my take.
Just wanted to introduce the possibility...should we get more participants in following years.

Well...Markus has had near a week now, and no objection has been posted, so, I believe we will go forward with the new plan.
As it is, per the original rules, to change the weighting, we only needed a majority, not a unanimous consensus.

That said, if Markus had had any serious objections, I would have attempted to address them.

The next Spreadsheet update will reflect the new weighting scenario.
The weighting will be updated along with each new update, weighting factors will be based on the individual's statistics at the time of the Update.

Updates will occur approximately twice a month, the dates of which shall be different...and unannounced...to prevent any person from sandbagging to get better stats.

Technically, of course, I COULD sandbag, since I will know...but if anyone notices my regular entering habits...they will see I enter a slew of bills at a time...usually every 4-6 weeks...and then nothing until the next slew of bills is to be entered.

SO....the next update will reflect the new weighting.

Moving discussion over here:

Kalisiin wrote:I'm working the stats, and this looks like Tony gets a total runaway because of the West Bonus...intended to give a boost to Dave. It is creating a runaway situation that we wanted to avoid. I'm reducing the West Bonus to 1.1 if there are no objections.

Huh? I'm not sure I agree. I think I counted 52 claimed hits for Tony (may now be 51). That means you need 4 hits to pass him under the current system. Giving Tony a 1.1 West coast factor gives him just a 2 tenths of a point lead.

dctim wrote:Moving discussion over here:

Kalisiin wrote:I'm working the stats, and this looks like Tony gets a total runaway because of the West Bonus...intended to give a boost to Dave. It is creating a runaway situation that we wanted to avoid. I'm reducing the West Bonus to 1.1 if there are no objections.

Huh? I'm not sure I agree. I think I counted 52 claimed hits for Tony (may now be 51). That means you need 4 hits to pass him under the current system. Giving Tony a 1.1 West coast factor gives him just a 2 tenths of a point lead.

Wrong. He has 51. Dade, FL was taken. And here's the scoop...
Old system - Tony factors = 1.0 for Year, 0.98 for Entered, 0.95 for Track Points, 1.0 for hit rate
Take those...1.0 x .98 x .95 x 1 = 0.931 - NOW...use a 1.4 West Bonus...he gets 1.3034 - WAY UNFAIR given his stats.

New way - 1.0 x .98 x .95 x 1 x 1.1 = 1.0241

Given that he joined before me...has WAY MORE bills entered than I do...and WAY more track points than I do...there is NO WAY IN HELL he should have weighting factors even higher than mine...and almost as high as Jackson...who really is a newbie!
Quite frankly, given his stats, he's DAMN LUCKY his weighting is as high as it is with the new system...look at where Markus is compared to Tony.
How could you justify Tony with a 1.30 weighting while Markus is at 0.84??

NOT RIGHT!!!

The West Bonus was INTENDED to give a boost to Dave...it would create a runaway for Tony now..his joining threw a monkey wrench into what was originally goodwill.

I think even tony would agree that it would be way unfair to everyone, given his claims thus far...which is the same as mine...his weighting should be close to mine...and it is. It's justifyably lower than mine as his stats are way higher.

Cross-Post wrote:EDIT: Justification for this - The old formula gave Tony a Weight factor of 1.3034...against Markus having 0.8471...seeing ads how Tony has so many bills entered, and so many hits...this could not be justified. I think the new adjustment is more fair to everyone.

This is perhaps where I am confused. I have Tony's Average weighting at 1.062, not 1.3034.
1.00 + 0.98 + 0.95 + 0.98 + 1.40 = 3.91 ... 5.31/5 = 1.062

For Markus, I also have different numbers:
0.90 + 0.98 + 0.98 + 0.95 + 1.00 = 4.81 ... 4.81/5 = 0.962

In regards to Dave, he is not a stamp-them-like-crazy kinda guy, so he reduces the number of hits he could be getting.

dctim wrote:This is perhaps where I am confused. I have Tony's Average weighting at 1.062, not 1.3034

Where do you come up with THAT?
Do the figures on a calculator!

In either case - THE WEIGHTING IS INTENDED TO BALANCE THINGS!!

No WAY should Tony be ahead of me in score...with the exact same number of claims.

dctim wrote:This is perhaps where I am confused. I have Tony's Average weighting at 1.062, not 1.3034.

1.00 + 0.98 + 0.95 + 0.98 + 1.40 = 3.91 ... 3.91/6 = 1.062

In regards to Dave, he is not a stamp-them-like-crazy kinda guy, so he reduces the number of hits he could be getting.

You are doing the weighting differently than I am, then.
I am doing 1.00 TIMES 0.98 TIMES 0.95 TIMES 0.98 TIMES 1.40 EQUALS 1.3034

Everyone's weighting was done that way.

Kalisiin wrote:

dctim wrote:This is perhaps where I am confused. I have Tony's Average weighting at 1.062, not 1.3034.

1.00 + 0.98 + 0.95 + 0.98 + 1.40 = 3.91 ... 3.91/6 = 1.062

In regards to Dave, he is not a stamp-them-like-crazy kinda guy, so he reduces the number of hits he could be getting.

You are doing the weighting differently than I am, then.
I am doing 1.00 TIMES 0.98 TIMES 0.95 TIMES 0.98 TIMES 1.40 EQUALS 1.3034

Everyone's weighting was done that way.

Interesting that you are doing it that way - it sure explains why we have different results.

Any particular reason you choose multiply over add?

No particular reason...let me work up the figures your way and see what we get...

OK, doing it your way we get these figures...
Kalisiin 1.028
Tony 1.046 / 0.986
Markus 0.968
DCTim 0.986
Jackson 1.062
Dave 1.076 / 1.01

I did Tony and Dave as a 1.4 versus 1.1 west Bonus.

I like your way better, Tim..of adding then dividing, rather than multiplying, because it addresses Markus' unfortunate situation.

HOWEVER, I still am going with a 1.1 West Bonus, because, sorry...tony has way better stats, and way more bills out there than I do...and has been here six months longer. Given that...there is no way he ought to be ahead of me with the exact same number of counties claimed...51.

Close, yes...but not ahead of me. Not unless he were to get another four or five counties...THEN it would be justified for him to be in the lead.
Remember what the purpose of the weighting was...to avoid a runaway, and to balance off people with huge stats against people with lesser stats...so that everyone would have a reasonable chance of winning.

It is not reasonable that Tony would be ahead of me FOR THE EXACT-SAME NUMBER OF COUNTIES!

Unless there is any major objection, I am going with Tim's formula...and a 1.1 West Bonus only.

This seems to balance things out better.

I'll post the example of this on the spreadsheet.

all done!!

The stats look much better and fairer now.
Have a look...
Everyone needs between a low of 47 and a high of 52 claims to get a 50 Score.
Everyone gets a score of range 48.4 to 53.1 for 50 claims.

Now THAT's some more realistic handicapping, all the way around!!

It fixes the problem of Markus being at a huge disadvantage...
It does address the fact that West Coast counties are bigger, thus giving a player a little disadvantage in their home state...
It does not create any runaway...
It addresses the fact that someone who has been on longer...and who has EXACTLY the same number of claims...SHOULD have a slightly lower score.

THIS is a formula that I think can work.

I agree -- the factors and standings look better.

I like your total for counties claimed - that's nice to see.

I'm gonna play around with my calculations - they seem to end with slightly different average ratios.

Glad you agree.

Seeing as how tony has 18,000 + Track Points and I have under 5,000 Track Points, there is NO WAY he should be ahead of me with the exact-same number of counties claimed!

By the way, in case anyone gets to thinking I'm fooling with the stats to give myself any advantage, know that I intend, if I win....to give the majority of it to Markus, in stamped bills, to support this site, and the awesome work he does for us!