Loose Deuces Video Poker: RTP – WTF : Getting a “Straight” Answer
I asked Google Claude to give me the actual return to player (RTP) for this version of Loose Deuces Video Poker. It was important because I was doing an ebook (on table game bonuses – and this video poker is eligible for the bonus).
I gave it the actual pay table, using maximum coins and asked “use the images to calculate the RTP is for this video poker game version – playing max coins and with optimal play : and check the rules to be sure there are no “bogus” rules that would offset the RTP”.
The first response
I was shocked a the first result. Google Claude said the RTP was both 95.35% and 98.94% – whoa. That was a real low figure, especially since ChatGpt gave me a RTP of 100.15% – it all looked legit – but certainly not consistent.



The 2nd Request: Clairification
For clarity the answer I simply asked a better more direct question “use the image to calculate the RTP is for this video poker game version DO NOT RELY ON MEMORY OR THIRD PARTY INFORMATION – DO THE ACTUAL MATH – playing max coins and with optimal play : and check the rules to be sure there are no “bogus” rules that would offset the RTP”
The initial “I am doing it” was encouraging : “I’ll calculate the RTP from scratch using the pay table shown in the image and combinatorial analysis of all possible 5-card draw poker hands with deuces wild”, but “The full combinatorial brute force is hitting a timeout. Let me use a smarter approach — exact frequency counting using known combinatorics for deuces wild, computed directly”, “The classifier is verified. Now let me run the actual RTP calculation — the full combinatorial enumeration is too slow in pure Python, so I’ll use numpy/caching tricks:”, “The deal-only numbers look correct. Now let me compute the proper RTP with optimal draw strategy using a smart grouping approach:”,
And then it just stopped
The 3rd and last Request
With a fresh chat on other day I give Claude a fresh request “using max coins and optimal play – what is the rtp of this loose deuces video poker game”
Based on that exact Loose Deuces (RTG-style) paytable you’re showing — and assuming:
- ✅ 5 coins (max bet)
- ✅ Perfect/optimal strategy
🎯 RTP (Return to Player)
- ≈ 99.07% RTP
- More precise calc versions: ≈ 99.072%
🧠 Why that number fits your screenshot
Your paytable matches the standard “500/5 Loose Deuces” structure:
- Royal Flush = 4000 (max coin bonus active)
- Four Deuces = 2500
- Straight Flush = 25
- Four of a Kind = 20
- Full House = 15
- Flush = 10
That exact structure is the common RTG Loose Deuces version, which is mathematically solved at ~99.07% with optimal play.
⚠️ Important nuance (this matters)
- This is NOT “full-pay” Loose Deuces
- Full-pay versions can exceed 100% RTP (player edge)
- Your version is slightly “trimmed” → casino edge ≈ 0.93%
💡 Bottom line
- Your game is actually quite strong (better than most slots)
- But:
- You need perfect strategy
- And max coins always (otherwise RTP drops hard)


