Wednesday, September 27, 2023

Proposal: Oddities of Odds

Withdrawn. Josh

Adminned at 29 Sep 2023 08:59:21 UTC

In the subrule “Giolitti” remove the bullet point beginning with “Odds”.

In the subrule “Odds”, replace the text “then Odds for all outcomes are the default value, which is 11:10.” with the following text:

then Odds for each Outcome are determined as follows:
* Determine the highest value of Track Record among all of the Participants for this Sporting Event, and set X for the Odds of every Outcome to this value except for “Draw”, if it exists.
* For each Outcome where one Participant wins, set Y for the Odds of that Outcome to the Track Record of that Participant for this Sport, to a minimum of 1.
* For the Outcome of “Draw”, set X for the Odds of that Outcome to the highest value of Track Record among all of the Participants, then subtract the Track Record of each remaining Participant from X, to a minimum value of 1 for X, and set Y to 1.

Any Punter or the Bookie should update the Odds of the Outcomes of a Sporting Event before 00:00 UTC on the day that Sporting Event occurs using the rules for Odds.

This isn’t a perfectly-accurate way of calculating Odds, but it might be close enough without being horribly complicated. It should be spreadsheet-friendly.

We could consider merging this with Track Ratio at some point, if it makes sense. I didn’t want to make any dependencies on it just yet.


Josh: he/they

27-09-2023 13:05:01 UTC

Good start.

The favourite is going to start with 1:1 odds; would be nice to see that come off the baseline a bit.


27-09-2023 14:17:17 UTC

Why get rid of the existing Giolitti formula?

JonathanDark: he/him

27-09-2023 14:35:06 UTC

I’m open to tweaks to the formula, if people have suggestions.

I removed the Odds formula for Giolitti because it didn’t seem to represent the actual track record of the winners. I’m willing to remove that line and leave the Giolitti rules alone if the rest of the proposal is otherwise acceptable


27-09-2023 15:24:22 UTC

I’d need to do some math to see under what circumstances the new numbers are reasonable - I already did that math for the current Giolitti odds.

This formula has an obvious issue with ClingBoom’s odds not adjusting for the extra participants, though, which makes ClimgBoom bets almost certainly terrible.

JonathanDark: he/him

27-09-2023 15:31:09 UTC

It’s probably worth defining specific Odds rules for ClingBoom, if that’s the outlier. I’d rather do that than make the overall formula more complicated to accommodate just one Sport.


27-09-2023 15:31:14 UTC

Also, a participant with a Track Record of 0 results in a divide-by-zero payout.

JonathanDark: he/him

27-09-2023 15:37:51 UTC

Good point, I just now added a minimum of 1 for Y.


27-09-2023 16:41:32 UTC

Since Track Record 0 is presumed a longer shot than Track Record 1, I’d double X in addition to setting Y to 1.


27-09-2023 21:35:12 UTC

For today’s Nomich match, this would give odds of:
6 : 6
6 : 6
1 : 1
So even odds all around and all worse than the current default odds.

For tomorrow’s Blogbots, it would give odds of
Ducktank Win 8:1
Hard Hat Win 8:8
Ducktank SD.  8:1
Hard Hat SD.  8:8
Draw       7:1

Where giving even odds to the favorite seems extremely generous, and Sudden Death odds are bogus anyway.

...not that the current odds are all that rational outside Giolitti, but this doesn’t seem to be much improvement. against

JonathanDark: he/him

27-09-2023 22:44:01 UTC

I welcome suggestions for improvements that don’t overcomplicate things or require simulations via coding. Simpler tweaks are better.

Maybe the favorite in each Sport gets X:X+Y odds rather than X:X, so that you get paid a bit less for betting on the favorite.

If both Participants have the same Track Record values, make it X+1:X odds for each, which is similar to but not exactly like the current odds.

I did struggle with Draw, especially since it can vary with each Sport. Some Sports can naturally lead to a Draw. With others, a Draw is highly unlikely. Now that the Method of Resolution is hidden, and with support building to simplify this Method into just a formula, it’s hard to say.

At the end of the day, the information we will have is the Track Record, which should probably be regenerated if a Proposal to reduce the Method of Resolution to a formula is enacted.

JonathanDark: he/him

27-09-2023 22:44:52 UTC

In any case, the current set of Odds needs improving, and I think it’s better to get incremental improvements than outright rejections with no follow-up suggestions.


28-09-2023 01:51:46 UTC

for Guy Wins outcomes, X = 3 plus total of other participants’ track records, and Y = own track record.
for Draw outcomes, X = total of all participants’ track records and Y = 3.

Three is not necessarily the right constant to use for the Draw’s effective Track Record.

Kevan: he/him

28-09-2023 09:05:28 UTC


Josh: he/they

28-09-2023 09:20:15 UTC

against Per Bucky. I think this is the right approach, just needs a bit of fine-tuning.

JonathanDark: he/him

28-09-2023 13:21:19 UTC

against Withdrawn, I’ll re-propose with Bucky’s suggestion