How to win at gambling without fail

The Apaṇṇakasutta (MN 60) subverts the imagery of gambling by reasoning towards logical certitude in a world of apparently random chances.

Regards the key term apaṇṇaka, the Critical Pali Dictionary says the etymology is unknown, and Cone has a question mark. CPD notes:

perhaps originally a term from the game at dice (the proper sense of which had been forgotten).

It is sometimes derived from pañha with the sense “unquestionable”. Most translators lean on this derivation, with such renderings as “incontrovertible”, “unquestionable”, etc.

However, I’ve always found this a little, well, questionable, perhaps because nowhere else in Pali does pañha appear as paṇṇa. Also, there seems to be no real cognate for it in Sanskrit, forcing later Buddhists to invent a hyper-Sanskritized form. Something always felt a bit post-hoc about this explanation.

The commentaries don’t, so far as I know, support this derivation. The commentary for MN 60 says:

Apaṇṇakoti aviraddho advejjhagāmī ekaṃsagāhiko
Apaṇṇaka means unfailing, an undivided path, categorical

The Apaṇṇakajātaka (Ja 1) offers a similar explanation:

Tattha, apaṇṇakan-ti ekaṁsikaṁ aviraddhaṁ niyyānikaṁ.
In this connection, apaṇṇaka means being sure, unfailing, leading to deliverance.

Now, if we look at words that do appear as paṇṇa in Pali, we find the number five (pañca), of which paṇṇa is a common form. To understand the significance of this, we’ll have to spend a little time at the casino.

Apaṇṇaka is closely connected with the gambling term kataggāha. Both feature in MN 60 and in SN 42.13. Further, apaṇṇaka describes the throw of a “gem”, apparently a kind of die, at AN 3.118:4.6 and AN 10.217:17.1. The extended metaphor the Buddha develops in all these cases is that, while it may appear that life is but a game of chance, it is possible, using the wise application of reason, to find a path that can succeed without fail.

Gambling was a major feature of Vedic culture, with the moving tale of Rig Veda 10.34 recounting the thrill and loss of the game. The gods were invoked to ensure success (Atharvaveda 7.109). Even kings bowed to the dice (Rig Veda 10.34.8), so that the Rājasūya consecration ceremony is secured with a (loaded) game of dice (Śatapatha Brāhmaṇa, 23). Numerous tales of kings losing their realms at dice, most notably Yudhiṣṭhira in the Mahābhārata, show that the power and danger of gambling was not merely symbolic.

Deatails of the classical Vedic game of chance are obscure and would have changed over time, but it seems that a large quantity of vibhītaka (“bedda”) nuts were cast in a hollow, from which players took a handful. If the number of nuts was divisible by four, it was said to be a “perfect” (kaṭa; see Rig Veda 1.132.1, etc.). This explanation is from the Jamison/Brereton translation of Rig Veda 10.34, relying on a 1986 German article by Harry Falk, Bruderschaft und Würfelspiel: Untersuchungen zur Entwicklunsgeschichte des vedischen Opfers.

The Pali expression kaṭaggāha, often translated as “winning throw”, therefore means “perfect (kaṭa) hand (gāha)”.

Its opposite is the kaliggāha, a “losing hand”. While a set of four dice was “perfect”, the fifth “losing” (kali) die was all-powerful, since just one extra die meant you lose everything (Rig Veda 10.34.2, Śatapatha Brāhmaṇa

Apaṇṇaka is therefore literally a “set without a fifth”, and metaphorically “unfailing”.

Of course if you really want to win at gambling “without fail”, that’s simple. Don’t play.


By the way, if you’re interested in a deep-dive into the logical methods of the sutta, check out Bronwyn Finnigan’s article, The Buddha’s Lucky Throw and Pascal’s Wager.

BuddhismPascalWager.pdf (480.3 KB)


Thanks for another interesting read! Obviously gambling was as popular in ancient times as it is today. Still, people are loosing their “entire kingdoms” nowadays. The methods are probably a bit more sophisticated.

I can’t find apaṇṇaka anywhere in the 24th Samyutta.


The Buddha’s path may be unfailing, but I sure am not. It should be SN 42.13. I’ve corrected the post, thanks.


Better luck now, thanks! :game_die:

1 Like
1 Like


This brings up a very vague (therefore not sure how accurate) memory from High school when we talked about ancient Egyptian hieroglyphs, and the teacher mentioned that numbers were only represented accurately up to 4: “1–2–3–4–many”.

1 Like

How does apaṇṇaka mean “(set) without a fifth” as pañca (five) is not paṇṇa in Pāli? Pañca can become pañña- (as in paññāsa i.e. Skt. pañcāśat “fifty”) but not paṇṇa-

There is the word paṇa in Sanskrit which means “a bet”, or “a wager” - so paṇya(ka) could possibly mean something that could either be won or lost as a result of the bet/wager (and apaṇyaka its opposite i.e. no possibility of loss). But I fail to see where there is any indication of the number five (either as cardinal or ordinal) in this word.

For what it’s worth, the Pali number for 50 is paṇṇāsa (at least for the Majjhima Nikāya sections), and paṇṇarasa is sometimes used for 15.

It’s not clear to me why we sometimes see the paṇṇa- form instead of the more expected pañca-

There is a Pali verb paṇati, risk, bet.
Perhaps apaṇṇaka can have the sense of ‘without risk’ , infallible.

It seems both variants (paññāsa & paṇṇāsa) exist in the canon - and they both correspond to Skt. pañcāśat (“fifty”) - and both variants are listed in the DPD as meaning ‘fifty’.

1 Like

It not clear to me how ‘paṇati’ can be related to 5 or 50, but I suppose it’s possible.

It may, instead, be related to the root ‘paṇ’ - ‘bargain’.
E.g. āpaṇa/ āpaṇika - shop/shopkeeper

The Pali verb paṇati is not related to pañca.

Paṇati is derived from the verb root ‘paṇ’ which is listed with the following meanings in Monier Williams dictionary (in sanskrit):

  • to honour, praise, Naigh. iii, 14 ;
  • to barter, purchase, buy, Br. ; MBh. ;
  • to negotiate, bargain, Āpast. ;
  • to bet, stake, lay a wager, play for (with gen. [e.g. prāṇā-nām, Bhaṭṭ. ; cf. Pāṇ. ii, 3, 57., Kāś. ] or acc. [e.g. kṛṣṇāṃ] MBh. );
  • to risk or hazard (as a battle), MBh. ;
  • to win anything (instr.) from (acc.), ib.
1 Like

Yes, which is pretty much what I wrote in my post above.