Pai Gow Gee Joon Tile Ranking Question

The Gee-Joon tiles used together form the Supreme Pair. When used separately they are semi-wild and counted as either 3 or 6 in non-par hands which are valued by the sum of tile pips. Should non-pair hands have the same numeric value the tie is broken by comparing the rank of the highest tile in each hand.

Tile ranks follow pair ranks with the exception of the Gee Joon tiles which lose their supreme ranking. According to the Wizard of Odds Pai Gow Tiles, Rules, 10, “the highest ‘gee joon’ pair are ranked lowest individually and will therefore never be a hand's high tile.” I have seen similar statements from other sources.

While doing some Pai Gow tiles analysis I can across a diagram in Resort World Catskills’ Introduction to Pai Gow Tiles document that I infer says that is not so exactly. I think there may be an exception. Just one.

The diagram shows the military/mixed tiles ranked by number of pips, just like the pairs. But among those tiles are the Gee-Joon-6 tile ranked after the Chop-Chit-7 tiles and before the Chop-Ng-5 tiles. The Gee-Joon-3 tile is shown as ranked last, after Chop-Ng-5 tiles. This strikes me as a logical way to account for the Gee-Joon tiles when they don't together form the Supreme Pair, allowing them to have some value rather than none.

That value wouldn’t be enough to matter, except in one case. That case is two hands each having a Chop-Ng tile and a Gee-Joon tile. Both Gee-Joon tiles would be valued at 3 and both hands would have a numeric value of 8. If the Gee-Joon tiles truly have no value the Chop-Ng tile comparison as high tiles results in a copy had the banker wins. But, if the ranking presented by Resorts World Catskills is in effect the hand with the Gee-Joon-6 tile wins.

Did Resort World Catskills get this wrong? Did most sources about tile ranking miss an obscure point that only matters for 1 in 35,960 comparisons? Comments, please. (And vote!)

The Wizard clarifies in this [cued] video...



...so the ranking might be correct but, since the 2-4 Gee Joon would be a three when paired with a Chop Ng, it mathematically never matters.

Edit: That is, if you declare your wild 2-4 to be a six, your declared hand's value becomes a one (5+6) and you'd lose to the eight (5+3). No tie at all.

Quote: BleedingChipsSlowly

Did Resort World Catskills get this wrong? Did most sources about tile ranking miss an obscure point that only matters for 1 in 35,960 comparisons? Comments, please. (And vote!)

No, they didn't get it wrong, but it's just because of the quirky way the New York State regulations are written, and which the Resorts World literature seems to have followed. It doesn't change any of the basic pai gow rules. If the player and banker both have Gee+5, the hands will be a copy and the Banker will win.

The relevant section is NYCCR 9-5324.50(a)(1)(5)(b)(4) which says:

"...(4) Notwithstanding the provisions of paragraph (3) of this subdivision, if the tiles that form the supreme pair are used separately, the numeric total of the 3 (1-2) may be counted as a 6 and the numeric total of the 6 (2-4) may be counted as a 3. When the 3 (1-2) is counted as 6, its individual ranking shall be fifteenth instead of seventeenth and when the 6 (2-4) is counted as 3, its individual ranking shall be seventeenth instead of fifteenth...."

So, whenever the Gee is counted as 6 it's NEVER the determining tile in hands of equal value. On the flip side, if the Gee is played with the 5 (the only relevant scenario for this ranking rule) it's always counted as a 3, and thus under the NY rule, it's ranked 17th, which is lower than the 5, and the only scenario where this would be an issue is where the other hand had a Gee + 5, in which case the Gee there would also be ranked 17th, and there would be a copy, which would go to the Banker.

Thank you Gialmere and UP84, your responses resolved my question. I have voted for the third option, and I will explain.

The statement I cited from the Wizard of Odds site is incorrect, and The Wizard himself explains in the video Gialmere provided, that the Gee Joon tiles do have a ranking value when not played as the Supreme Pair. So the first voting selection gives the right result for the hands valuation but that result is predicated on a false statement.

The second voting selection is also wrong. The bolded text UP84 quoted from New York gaming regulations provides the clear and concise rule governing how Gee Joon tiles are ranked. The Resorts World Catskills diagram is, at best, misleading absent the rule explaining that the ranking assigned to each of the Gee Joon tiles is tied to the value declared for arriving at the hand’s numeric value.

Thanks again, Gialmere and UP84, for shining a light on this issue. But one more thing…

Quote: Gialmere

...
Edit: That is, if you declare your wild 2-4 to be a six, your declared hand's value becomes a one (5+6) and you'd lose to the eight (5+3). No tie at all.

… but if both the dealer and player declared their respective Gee Joon ties to be counted as six, then the hands would be numerically equivalent. Then the Gee Joon tiles would rank highest in each hand and be used to resolve the tie. Of course, they would rank equally and the hand would still be a copy. So it might be theoretically possible for the Gee Joon tiles to be used as the highest ranking tiles.

I’ll be getting my coat…

I've always hoped to get a gee-joon+5 tie with the dealer with the 2-4 tile in my hand so I could pull out the game rules and argue that I should win the ranking, but it's never happened to me.

If this does happen I would suggest the casino follow its own rules as printed in its game guide and give the tie to the player with the 2-4.