Project Update (09/08/2026)
- Red Cards: my model for some reason was not counting them into decision making. Root causes identified, and fixed.
- Odds: Today my odds provider had delays, which made some alerts come out with slightly inflated/calculated edges + generated alerts (too late). I need to think on guardrails for this, and will fix it. 🧠
Thanks for your continuous support.
- Red Cards: my model for some reason was not counting them into decision making. Root causes identified, and fixed.
- Odds: Today my odds provider had delays, which made some alerts come out with slightly inflated/calculated edges + generated alerts (too late). I need to think on guardrails for this, and will fix it. 🧠
Thanks for your continuous support.
👍5
A clear example of the Odds issue today. Cuiaba scored 1-0, so the real live odd of Cuiaba to win was 1.65.
But, the Odds feed stayed stale at 2.75, so the Sigma Predict model evaluated based on that and found "big edge" (of course).
Such an edge, not only triggers bad alerts for users, but also takes more stake units from bankroll. So overall bad for everyone.
This is an Odds provider issue not Sigma Predict issue. But I will implement protection mechanism for this. 🤞🏻
But, the Odds feed stayed stale at 2.75, so the Sigma Predict model evaluated based on that and found "big edge" (of course).
Such an edge, not only triggers bad alerts for users, but also takes more stake units from bankroll. So overall bad for everyone.
This is an Odds provider issue not Sigma Predict issue. But I will implement protection mechanism for this. 🤞🏻
🔥2
Sigma Predict is an autonomous sports prediction engine, that fires high probability alerts when a market edge is detected.
Core values:
Full Transparency & Continuous Improvement
We automatically keep track of all alerts for analysis and model improvements, but also to show to public and subscribers. No edits, no faking, pure transparency.
Results are good, we show. Results are bad, we also show and improve.
FREE TRIAL use: @SigmaPredictBot
Our main website is: sigmapredict.com
Support: support@sigmapredict.com
And our oficial social networks are:
- Instagram
- Facebook
- X/Twitter
- YouTube
Core values:
Full Transparency & Continuous Improvement
We automatically keep track of all alerts for analysis and model improvements, but also to show to public and subscribers. No edits, no faking, pure transparency.
Results are good, we show. Results are bad, we also show and improve.
FREE TRIAL use: @SigmaPredictBot
Our main website is: sigmapredict.com
Support: support@sigmapredict.com
And our oficial social networks are:
- X/Twitter
- YouTube
Odds Topic Update
Clearly, I need to change my Odds provider. Not only is it freezing sometimes with stale odds during match, but also injecting pre-match odds during live. This is non-sense...
Really sorry for this, but I'm trying hard to fix this or at least protect the system. But everyday a new Odds issue.
I will keep you all updated.
Clearly, I need to change my Odds provider. Not only is it freezing sometimes with stale odds during match, but also injecting pre-match odds during live. This is non-sense...
Really sorry for this, but I'm trying hard to fix this or at least protect the system. But everyday a new Odds issue.
I will keep you all updated.
👍2
Daily Project Update
I have implemented two protection mechanisms due to external Odds provider issue. And Red Card protection in model.
1) Pre-match odds injection: Yesterday I have noticed that my provider was injection during live matches, some pre-match odds. This is non-sense, but I implemented a protection fix now.
2) Stale odds: external odds provider freezed odds (stopped updating) in some cases. Which made odds stay higher for the model, but in reality they where lower. Not many cases, still, I have fixed it.
3) Red Cards: due to an unknown reason, my model stopped considering red cards into calculations. Fixed now.
I will monitor everything closely in the next few days, to see system stability. Until then
All active Trial accounts via @SigmaPredictBot will be whitelisted until further notice, instead of 10 days only.
Thanks all for your understanding and support! I'm pushing hard to make the best system ever created. 🧠
I have implemented two protection mechanisms due to external Odds provider issue. And Red Card protection in model.
1) Pre-match odds injection: Yesterday I have noticed that my provider was injection during live matches, some pre-match odds. This is non-sense, but I implemented a protection fix now.
2) Stale odds: external odds provider freezed odds (stopped updating) in some cases. Which made odds stay higher for the model, but in reality they where lower. Not many cases, still, I have fixed it.
3) Red Cards: due to an unknown reason, my model stopped considering red cards into calculations. Fixed now.
I will monitor everything closely in the next few days, to see system stability. Until then
All active Trial accounts via @SigmaPredictBot will be whitelisted until further notice, instead of 10 days only.
Thanks all for your understanding and support! I'm pushing hard to make the best system ever created. 🧠
👍3
Please open Telegram to view this post
VIEW IN TELEGRAM
Daily Project Update
- Odds are stable, engine more robust 🙏🏻
I made a regression analysis on previous results with the odds problems and the impact was ~30units. I reconciliated the public ledger to reflect this, for more realistic results. Overall ROI dropped 2.7%, sits now around 13%.
- Limited the total exposure to max. 5 units per match. More conservative engine, but saves on unnecessary over exposures.
- Added Saudi Arabia League. Tomorrow, engine will already monitor the matches there.
Trial whitelist still active, at no cost: @SigmaPredictBot
- Odds are stable, engine more robust 🙏🏻
I made a regression analysis on previous results with the odds problems and the impact was ~30units. I reconciliated the public ledger to reflect this, for more realistic results. Overall ROI dropped 2.7%, sits now around 13%.
- Limited the total exposure to max. 5 units per match. More conservative engine, but saves on unnecessary over exposures.
- Added Saudi Arabia League. Tomorrow, engine will already monitor the matches there.
Trial whitelist still active, at no cost: @SigmaPredictBot
👍2
This afternoon, between two losses, 10 winners. These kind of winning strikes are the ones worth waiting for at Sigma Predict.
Are you in or out? Get in: @SigmaPredictBot
Are you in or out? Get in: @SigmaPredictBot
👍2
From today on, Paid Subscribers will have access to:
- Private Telegram live alerts channel
- Pre-match analysis dashboard with all the baselines, analysis, context and suggested markets (same info the live models use as a base)
- Results dashboard with all previous results from the live telegram signals, with filtering options to make your own analysis.
This project's core value is: TRANSPARENCY
I'm building what I think will help a lot of people on a daily basis, the same way it helps me.
If you are not yet a paid user, now is the best moment to do so. Start monthly here.
- Private Telegram live alerts channel
- Pre-match analysis dashboard with all the baselines, analysis, context and suggested markets (same info the live models use as a base)
- Results dashboard with all previous results from the live telegram signals, with filtering options to make your own analysis.
This project's core value is: TRANSPARENCY
I'm building what I think will help a lot of people on a daily basis, the same way it helps me.
If you are not yet a paid user, now is the best moment to do so. Start monthly here.
🔥2👍1