Today we are announcing the introduction of a change to the algorithm that calculates our Tipster Ratings at Pyckio. We know there will be tipsters who will be harmed by these changes and therefore won’t like this modification. However, our intention is to make our tipster evaluation as representative as possible of the quality and skills of the tipsters themselves. Moreover, our new Rating has been designed taking into account our future Sports Betting Fund that will feed on Pyckio tipsters/analysts to place its bets. In fact, our new Rating includes bet liquidity as a factor, something vitally important to manage a portfolio that has sports wagers as an asset.
One of the motivations that has made us change our rating is the aggressive and chaotic staking policies of some tipsters, carried out with the intention of artificially manipulating our Rating. We can see this with the following two examples:
Case 1: One Tipster makes 400 bets with an average stake of 2. He reaches a 3% yield. Afterwards he suddenly starts making stake 10 bets. Each of these new set of bets is going to weigh 5 times more on his aggregate yield than the previous ones. If the tipster experiences a good run in a small amount of bets – something which is quite possible – he can raise his yield from 3% to 10%.
Case 2: One tipster reaches the 500 bets figure with an average stake of 8 and a 15% yield. To “protect” this yield he starts making stakes 1 or 2. In this way, even though the tipster’s results with his new staking policy aren’t positive, his 15% yield isn’t going to drop too much as the losses of the new bets have a small impact on the global results. Therefore, the tipster can reach 1000 bets with a relative high yield, even though their results in this second stage have been poor.
In 2015 we introduced the “Level Stakes Yield” ratio in our algorithm to compute not only the tipster real stakes but also their flat state ones. However, this does not completely correct the practices of some tipsters who try to manipulate their Rating with their stakes. Moreover, in practice, most bettors do take into account the odds to decide their stake, as the Kelly method suggests. That is, even though a Tipster always does level stakes and therefore recommends the same stake at odds of 1.25 or 8.0, in general bettors don’t do that; the higher the odds the lower the stake and vice versa, in a way that the profit target of every bet doesn’t vary too much. For that reason, in our new Rating system, different stake strategies are taken into account in a way that manipulation is virtually impossible. It doesn’t mean that you cannot eventually do higher stakes than average when you’ve got confidence in a bet but that a staking policy drawn to manipulate our rating will penalise you.
Aditionally our new rating includes the following factors:
- The number of picks submitted by the Tipster now has a slightly higher weight than before.
- The time the tipster has been tipping. That is, between 2 tipsters with the same yield and number of picks, the tipster who has built his record in six months will have a higher rating than the one who did it in one or two. Our data confirms to us that the probability of a tipster who has achieved good results being due to luck is higher if the record has been obtained in a short period of time. This is logical as if you don’t filter the number of picks you make it more difficult to find value odds.
- The liquidity of the matches where the tipster tips. All else being equal, the higher the liquidity the higher the rating, as the odds drop for the punter is lower for more liquid fixtures and vice versa. To measure liquidity we factor the money Pinnacle lets us bet at the moment the tipster submits his pick. The liquidity offered by Pinnacle depends on both the type of market and the time until the kick off. It could happen that a Premier League match has low liquidity because the tipster sent his pick right after the odds have gone out or a Handball game having high liquidity because the pick was sent a couple of hours before the start.
To sum up, these are the changes that affect our Ratings:
1. Results under different staking strategies are considered.
2. The number of picks gains importance.
3. We take into account the time the tipster has been tipping.
4. Liquidity of the picks submitted is considered.
We’ve carried out many simulations that prove that our new rating better predicts the future results of our tipsters.
In any case, we’ll always require min. 500 picks and a min. 4.5% Yield LS to become a PRO Tipster.
According to these guidelines the perfect tipster would be one with a high Yield, who has submitted a high number of picks during a longer time frame, who follows sound staking management and who provides high liquidity tips. If you are a tipster at Pyckio, keep on focusing on finding value odds and do not get crazy with your stakes.