Spin & Go Max Experiment

I really don't want to turn this into a whinging thread, but once again:

According to Nash, that was a proper shove. His call was ludicrous.

I would be nice if these things started to even out soon!
VS

Good luck! I tried some of those on stream, also 1$. Super random and spewy

Spewy is too mild a word!
Problem is, I'm the one spewing!

I'll follow your stream for a few hours!

Cheers,
VS

Originally posted by VorpalF2F
Spewy is too mild a word!
Problem is, I'm the one spewing!

I actually like the idea of random table sizes, and I don't care about the extra randomness in the prize pool. It's just that the hand limit sucks, especially with the high rake.

Originally posted by VorpalF2F
According to Nash, that was a proper shove. His call was ludicrous.

Why did you set the 80% / 20% payout structure? Remember that you need to input the EV of the 1st place prize (the arithmetic mean of the 3 possible prizes), not the biggest of the 3 options. When a Spin & Go Max pays 2 places, the distribution of the expected prize pool is usually between 62.5% / 37.5% and 75% / 25% in favour of the winner, most often either 66.7% / 33.3% or 70% / 30%, but never 80% / 20%.

It does look like you're not going to beat this game if you don't even know how which payouts to put in the calculator.

That said, the shove was correct anyway and BTN's call was beyond terrible.

It’s a tricky format for me. I don’t like it it’s random that the winner gets a good price as well.
This way 100 games are nothing to have an indication if you are beating the game or not. You most likely need more then 50k games or so and that’s almost impossible to reach in a year I think.

Cheers,
SDK1987

Originally posted by SDK1987
This way 100 games are nothing to have an indication if you are beating the game or not. You most likely need more then 50k games or so and that’s almost impossible to reach in a year I think.

The estimation of the EV ROI is trickier here than in usual Spins, but not impossible. I'll have to do some SQL programming sometime in the future to get better precision than the one given by the indicators that can be calculated within PT4 itself (the chip EVs filtered by the number of entrants and the payout structure, which estimate the probability of finishing 1st, plus the actual frequencies of 2nd, 3rd etc. finishes).

Still, I wouldn't be able to judge by 100 games anyway. By 1K games - maybe.

Originally posted by tonypmm
Why did you set the 80% / 20% payout structure? Remember that you need to input the EV of the 1st place prize (the arithmetic mean of the 3 possible prizes), not the biggest of the 3 options. When a Spin & Go Max pays 2 places, the distribution of the expected prize pool is usually between 62.5% / 37.5% and 75% / 25% in favour of the winner, most often either 66.7% / 33.3% or 70% / 30%, but never 80% / 20%.

It does look like you're not going to beat this game if you don't even know how which payouts to put in the calculator.

Fair point, tonypmm,
I used it because that was the prize that was eventually won. I also ran the numbers using the lowest payout for comparison to see if it made any difference, and as you point out, it did not.

Since I made the initial post, my numbers have come way up, but still a long way to go. All you can do in these things is to watch for opportunities and pounce. In any case, this is what I thought I was doing, but the variance is huge.

I'll ride out the 100 games, but I don't think I'll play these often. I only play the Spingo format when I'm bored, or have only a short time to play.

I'll post the selection states a bit later this morning.
Right now, I'm watching w34z3l's stream.

Cheers,
VS

Originally posted by VorpalF2F
All you can do in these things is to watch for opportunities and pounce. In any case, this is what I thought I was doing, but the variance is huge.

There's little profit in poker without variance because many weak players (most of them, I guess) are drawn back into the game by the luck factor that blindfolds them if they run good - makes them think they're stronger than they in fact are.

However, some of the casual players don't like the swings, want to play postflop a lot, and both I and, fortunately, poker rooms understand their needs well, so I think that low-variance profit opportunities will be coming up as well.

As of now, I think that 888's Flopomania is such an opportunity because the amount that one can win / lose is capped (in the push/fold version), and the postflop play is nontrivial enough because everyone sees the flop with a 100% range. I'd gladly start learning it if it ran at higher stakes than $1 ante, which is roughly equivalent to $7 spins in terms of typical swings and perhaps profit. But as it is, I think my earnings will be potentially bigger in the long run if I become an expert at Spingo Max instead. If Stars introduces a jackpot version of Flopomania (like Natural8's All In or Fold Omaha, but with lower rake or better marketing), I'll consider learning it too - like Xanham, I don't mind the variance as long as the EV is definitely positive; just right now, it seems that there's too little demand for this game format except at low stakes.

Power Up is a rather low-variance game too, and very complex, just I'm afraid that it will be unpopular at high stakes because video gamers tend to be younger and have less spare cash to play with.

Things are improving.
80% of the way through, and I've picked up quite a lot. ITM is now a bit more respectable, but ROI still sucks.

I changed all the numbers to percentages, and added a line showing what I selected, and a line showing what I "would have" received had I always chosen the Max value. Had I done so, I would have a +ve ROI.

Since -- as tonypmm correctly points out, one should use the average of the 3 possible prizes, I included my results compared to the average. Note that those two lines are in $

Note that sample size makes this little more than an amusing exercise.

                   Max       Min       Med    Picked  
         Left:    40.7      38.3      21.0      30.9  
        Right:    33.3      30.9      35.8      21.0  
       Centre:    24.7      29.6      45.7      46.9  
     Selected:    29.6      42.0      28.4            
         Mine:    28.6      42.9      28.6            
                                               Total  
  Δ EV vs Max:            -23.50    -16.50    -40.00  
  Δ EV vs Avg:    8.50     -9.00     -4.50     -5.00  

Cheers,
VS

Notes:
 The "Left", "Right" & "Centre" rows show the percentage of time that column contained the Max, Min, and Middle value, alone with how often that column was selected by winners.
 The "Selected" and "Mine" rows show how often the Max, Min and Mid values were chosen by everyone, and "Mine" shows by me alone.

I just tried a couple of these and I must say that I enjoy this format. I wouldn't start grinding it but these are fun to play, at least in the evening after one or two beers.

Btw when you pick a prize is is totally random? Always 1/3 chance to pick the max value? Because I feel like I have been picking the largest prize a lot...

Originally posted by la55i
Btw when you pick a prize is is totally random? Always 1/3 chance to pick the max value? Because I feel like I have been picking the largest prize a lot...

It looks like I've found the one who's been stealing my luck

The chance is always 1/3, otherwise (in absence of exact odds stated) the gambling commission wouldn't allow to advertise the prize picking as 'random'. But for the sake of entertainment, I use the second hand of the system clock to decide which prize to pick

Quite often, the "Min" and "Mid" prizes are the same.
Much less often, the "Max" and "Mid" prizes are the same (I've only seen this once is 80-odd games)

VS

Originally posted by tonypmm

Originally posted by la55i
Btw when you pick a prize is is totally random? Always 1/3 chance to pick the max value? Because I feel like I have been picking the largest prize a lot...

It looks like I've found the one who's been stealing my luck

I use the second hand of the system clock to decide which prize to pick

Just always pick the middle one, has been working for me

That’s good to know, but then you will be wrong a decent amount of the time as well.

I guess I have just been very lucky I have a small sample so that is probably variance

Originally posted by la55i
Just always pick the middle one, has been working for me
Check the table. A lot of people pick the centre.
The centre has held the median value most of the time (so far -- sample size is really small)

VS

I won 4 spins today.. My "click on the middle" strategy seems to be working so far

Am I very lucky or what is this?
I already felt that I had picked the biggest pot quite many times and now today 4 times in a row.. The probablity of that is about 1%?

Well Played.
I managed 2 wins today out of 4 games -- one of them with a $15 max payout too, which would have put my ROI on the plus side for the first time since I started this.
Unfortunately, I picked the min both times.

sigh,
VS