Suppose you were to play Ronnie O'Sullivan in a frame of snooker. You would lose - you know this before you start; the only question being by how many points. The game simply requires too much sustained excellence - potting accuracy, cue-ball control, safety-play precision - for you to even contemplate an upset. Were you to play Ronnie at pool, however, and broke off by sinking one ball and spreading the rest, then provided you held your nerve, the chances are you could beat the greatest cueman the world has ever seen. Compression - potting eight consecutive balls on a table a quarter the size of a competition snooker table - provides opportunity.

Does the same logic apply to cricket: the longer the game, the less chance of a giant-killing? Does T20 carry the greatest chance of a potential upset, followed by List A (or ODI), followed by first-class cricket (or Tests)?

It was a discussion that surfaced in 2014, when Michael Vaughan floated the idea of an FA Cup-style domestic T20 competition - straight knockout, no seeding to protect the big clubs, with Minor Counties sides and others invited - to run on free-to-air TV alongside the existing NatWest Blast round-robin format. Of course, cricket's thralldom to finance and need for guaranteed fixture lists rendered the idea a non-starter, but it was interesting to ponder whether such a tournament would provide a greater chance of an upset - Vaughan said "5%" - than in the minor counties' 41-year participation in the Gillette Cup and its successors, which produced ten giant-killings in 336 matches. (In the Benson & Hedges Cup, the Minor Counties representative XI won six games out of 139.)

Does a shorter game really lower the odds, or on the contrary, has T20 engendered such a degree of specialisation that it would no longer be possible for the good amateur team - even bolstered by an overseas player or two, and playing on the poor club wickets on which most of those giant-killings took place - to contemplate an upset? Could it be that the skill set possessed by decent bowlers just below first-class level - shaping the red ball away, good accuracy, pace in the 75-80mph range - translate to white-ball T20 fodder? Is it only in the professional environment that players can hone the yorkers, slower balls, reverse sweeps, ramps and suchlike that are integral parts of the T20 repertoire nowadays?

What if we scale up to international level, to battles between the "lesser" teams (who might in any case be fully professional, as with Ireland) and established powers - do the same principles hold, even with better pitches and more evenly balanced attacks? Or, to ask a closely related question: which format most lends itself to Associates being competitive with Full Members, thus encouraging and nurturing the spread of the game?

Perhaps the recent self-interested reluctance of Full Members either to imbue Test cricket with more context in the form of the two-divisional structure, or to expand the game by providing more opportunity for Associates to participate in ICC global events - the next 50-over World Cup has been contracted from 14 teams to ten, and the format of the next World T20 is under discussion - renders this moot. Meanwhile Ireland's and Afghanistan's recent results provide ammunition for both the expansionist and pro-streamlining camps: the latter have won an ODI in Dhaka, beaten Zimbabwe 3-2 in ODI series and 2-0 in T20s both home (Sharjah) and away, and were, of course, the only team to beat West Indies in the recent T20 World Cup. Ireland, meanwhile, have suffered very heavy ODI defeats to Australia, South Africa and Pakistan in recent weeks, and slipped to 15th in the T20 rankings. Yet even if the question is increasingly hypothetical, it's still worth considering which format offers the best chance of an upset.

It is not inconceivable that an Associate that produced one all-time great bowler (Richard Hadlee and Muttiah Muralitharan carried New Zealand and Sri Lanka at times), backed by solid batting and disciplined support bowling, might compete in Test matches. Zimbabwe approached their first years as a Test-playing country by playing conservative cricket, hoping that taking the game deep would pile the pressure on opponents who were always expected to beat them.

Restricting ourselves to the two white-ball formats, conventional wisdom would suggest both that the 50-over game gives a team more chances to recover from a wobble than T20, its greater ebb-and-flow potentially beneficial to an underdog, and that victory in the latter only requires one or two players - or fewer than in 50-over cricket, at least - to perform well.

It is logical that as the scores rise in ODI cricket, close games will be less likely, period, let alone between teams unevenly matched on paper. No room for spoilers and grinders here - at least, not on good pitches

Evidence from World Cups in the two formats doesn't make an unambiguous case either way - and there are many different kinds of giant-killing, of course: low-ranked Full Member beats high-ranked Full Member; Associate beats low-ranked Full Member; Associate beats high-ranked Full Member in dead rubber; and the shockoholics' favourite, Associate beats high-ranked Full Member in "live" game.

There have been several victories by Associates over the high-ranked Full Members in the World Cup. Sri Lanka shocking India in 1979, Zimbabwe downing Australia in 1983 and England in 1992 (in a dead rubber), and Bangladesh beating Pakistan in 1999 all presaged their Full-Member status. Kenya beat West Indies in 1996, as well as Sri Lanka (and Bangladesh) in 2003. Ireland have, over the last three tournaments, beaten Pakistan, England and West Indies.

As for the six World T20s, a total of 24 games pitting Associates against the "Big Eight" throws up only three major upsets: Afghanistan's victory over West Indies, and Netherlands' two wins over England. Perhaps this reflects the fact that T20 is cricket's "state of the art", the place where the resources and systems of cricket's moneyed elite should be rendered swashbucklingly, bamboozlingly manifest.

And yet, this convergence of imagination, technology, bowler-punishing playing regulations and intensive training increasingly pervades the 50-over game. The two formats are becoming largely indistinguishable, as Eoin Morgan acknowledged in the wake of England's record-breaking 444. The knowledge that, in good batting conditions, 275 just won't be competitive pushes sides to recalibrate their risk-reward calculations, with the odd crash-and-burn 180 from going too hard, too early more than worth it for posting 300-plus and being competitive. It is this mentality as much as administrative tweaking that has all but killed the Boring Middle Overs.

Nevertheless, it is logical that as the scores rise in ODI cricket, close games will be less likely, period, let alone between teams unevenly matched on paper. No room for spoilers and grinders here - at least, not on good pitches. In addition, five in the circle until over 41, plus the Associates' general lack of bowling weaponry allows highly trained 360-degree batsmen longer to assert their supremacy. The chasm between amateurs and pros that T20 has opened up in English domestic cricket could be replicated by the "T20-fication" of the longer form of limited-overs cricket, creating a parallel chasm between Associates and Full Members.

And if competitive advantage in T20 is, arguably, to be found less in top-line bowling - neutered as it is by leg-side wide rules, bouncer rules, effective boundary size and so forth - and more in power-hitting capacity, then does the greater likelihood of a guy playing out of his skin for 25 balls than for 125 mean that, after all, T20 offers the better chance of an upset?

Scott Oliver tweets here