Moving on to the “guts” of the issue at hand. First off, my philosophy regarding combat odds and invasions – these considerations will effect all of the following. As an Allied player I would consider a 3:1 flat [i.e. no DRMs] attack as a sure thing. Regardless of the 1 in 6 chance for the devastating and disastrous AS on a die roll of one, I would always run this risk to grab a port.
As an Allied player I would consider a 2:1 flat attack as VERY risky, considering that it has a 50% chance of ending in disaster, on a dr of one through three – this is an attack that I might make, if I felt I had to, but 50/50 is not the sorts of odds I like when facing isolated losses [and note well that virtually ALL losses occurring during amphibious invasions will be, for the Allies, isolated losses. So given a chance to get ashore at 3:1 or better I’ll grab that opportunity, while having to risk 2:1 attacks will give me great pause.
As an Axis player I want to prevent ANY 3:1 attacks on a port during an invasion. I accept the risk of “only” defending with enough strength to prevent 3:1’s, and not 2:1’s, because I don’t believe it is remotely possible to defend EVERY port 2:1’s – and I believe it’s better and more economical to defend as many ports against 3:1 attacks, rather than defend some ports against 2:1’s and thereby leave some other ports [in the same basic geographical location] vulnerable to a greater than 3:1 attack.
Consider, Sicily has ten ports – thus to defend Sicily against an Allied invasion the Axis must prevent 3:1 attacks at ALL ten of these ports, lest the Allies guess wisely and get ashore at some under-defended port on “the cheap”. This would be a disaster, but for the Axis, not the Allies. Some simple math; you need 17 defense strength to prevent a 3:1 attack on a Med port during the Jul-Aug time period. A good Allied attack [not the best, but good] can generate 48 att str, for a US invasion that’s a 4-8 Inf III, 2x 3-8 Inf III, 2x 3-8 Cbt Eng III and 2x 1-8 Mar Cmmdo II. That’s 16 att str quartered [=4] and 2 att str halved [=1] for a total of 5 att str from ground units. NGS can provide a maximum of 25 additional Art str pts [4 per RE of non-Art units [6x 4 = 24] plus one additional pt which is rounded down to one pt, or a total of 25 att str. We now have 30 att str total.
With 6 REs participating in the attack, we can use up to 6 air units to assist the attack as GS. Our best air units right now have 3 TBFs each, so 6 REs worth of 3 TBF air units is 18 factors of GS which we can add to our attack – thus we now have 5 ground + 25 NGS + 18 TBFs of GS, for a total combat strength of 48. Given our att str of 48, any Axis position held by 16 or less DS [Defence Strength] can be attacked at 3:1 at least. Defending with forces which allow an Allied 3:1 attack upon landing means you are gambling your entire defence upon a 1 in 6 chance for an AS on a dr of one – some Axis players might take that gamble, I wouldn’t. Even the Allies’ “Amphibious Killer Wad” stack of all CW Mar Cmdos with a ground strength against non-Major city, non-coastal cliff hexes of 8 att str + 25 NGS + 18 GS or 51 total att str can still only muster a 3:1 vs a hex defended by 17 DS. Given that Palermo can be defended with only 13 DS [because, as a Major City it halves again all non-Eng, non-Art att str and GS], we can easily see that the Axis need 166 DS just to defend the ten port hexes – 9 ports requiring 17 DS [153 DS] and one port requiring 13 DS for a total of 157 DS.
Now, one could argue that Messina, because of it’s location, might be safe with a few fewer DS, but if our intention as the Axis is to defend Sicily, then we have to defend Sicily and that means accepting that the Allies will accept a 3:1 attack to get a port and thus we have to defend Sicily with enough strength to force the Allies to risk a 2:1 attack, at every port.
Of course, the above all hinges upon the Allies being forced to attack their desired port/s from a single hex, i.e. the port hex itself, where they disembark. If they can attack from even just two hexes, our required port defence strength is significantly increased. Assuming two identical US stacks as outlined above, the Allies can make a two hex frontage attack with 10 ground str, 49 NGS [6x 4 x 2, + the one additional Art str pt] and 2x 18 GS  for a total of 95 att str. Thus, against a two hex frontage invasion, the Axis would require 32 DS to prevent a 3:1 attack.
If you allow an attack on a port hex to be mounted from three hexes, then the Allies can throw in at least another 4.5 ground str, 24 more pts of NGS and another 18 TBFs of GS – now the Allies muster an att str of 141.5 – enough that it would require 48 Axis DS to prevent a 3:1…. and on and on. The lesson is simple – if you want to have ANY realistic chance to prevent a 3:1 attack on a port, you must limit the attack frontage to that single hex. Which the Axis can do. Around those 10 ports there are 23 hexes which are adjacent to one or more ports AND are allowable hexes for airborne and/or amphibious landings. If we put enough DS into each of these hexes, such that they are NOT susceptible to airborne or amphibious overrun, then we can force any Allied landing in these hexes to attack ONLY that hex itself, and be unavailable to assist any attack into a hex adjacent to the hex the Allied units landed in.
Now we know that the best possible Allied amphibious overrun [at this stage of the campaign] is 4 att str – this is the “Killer Wad” of 4x 3-8 Mar Cmdo X and 4x 1-X Mar Cmdo II. That’s 16 att str, halved for invading to 8 att str and halved again for lack of support [note well that NGS can NOT be fired during a movement phase, when overruns of any type occur and amphibious landings have to occur], thus 4 att str. So against an amphibious overrun, even an unsupported 1 DS unit, worth .5 DS, is strong enough to prevent an amphibious OV.
We also know that the best possible airborne OV would consist of 1x 4-5 Gldr X, a 4-5 Para III, a 3-5 Para X, a 3-5 Para III, a 2-5 Para X and a 2-5 Gldr III. That’s a total of 18 att str, all halved for lack of support, thus 9 att str pts IF [and this is a big IF when landing atop enemy units] not one of the airborne units is disrupted, or worse. But even if all airborne units land safely, they can still only generate 9 att str, which means 1 DS is sufficient to prevent the airborne OV. So either one supported DS or two unsupported DS is sufficient to prevent the airborne OV.
So, 23 hexes must be defended with at least 1 DS each to prevent a multiple hex attack against any of our ports. In a perfect world that would be 23x 1 supported DS units – however the Axis don’t have 23 self-supported 1 DS units available on or near Sicily. Being generous to the Axis, however, let’s stipulate that they can defend all of these 23 hexes with only 33 DS worth of units [since at least some hexes will require two unsupported DS…]. Thus, to minimally defend Sicily against an Allied invasion which we wish to prevent, we will require, at a minimum, 166 DS to protect the ports themselves from single hex attacks of 3:1 and require another 33 DS to set up an anti-airborne and anti-amphibious screen around each port to insure that no port on Sicily can be attacked from more than one hex.
166 + 33 is 199 DS. The Axis player, it should be noted, has a maximum total of Axis initial set-up units on Sicily worth 34 DS [the 1-2-10 PzGr II having converted to a 5-10 PzGr X] for the Germans, and 53 DS for the Italians for a total DS of 87 pts. [and note that again I am being generous, since I am assuming that ALL of the unsupported units are counted as being supported, even though that is highly unlikely.] Therefor, the Axis are short a minimum of 199 – 87 or 112 DS worth of units to defend Sicily adequately. Given that it is virtually impossible for the Axis to get an additional 112 DS to Sicily [recall the discussion of movement limitations above] by the beginning of the Allied Jul I player turn, it follows that it is virtually impossible for the Axis to adequately defend Sicily on the first turn of the game – if ever.
However, the sad news for the Axis doesn’t end here…oh no. To defend Sicily adequately, we also have to prevent it becoming cut off and isolated from Axis sources of supply – moreover we have to protect our reinforcement and escape route to and from Sicily. This requires, at an absolute minimum that we defend the two “toe” ports and Crotone [from which the Allies can supply an advance cutting off the entire “toe”]. That’s an additional 16 DS per three ports, or 48 DS and an additional four hexes adjacent to one or more of these ports, or 4 DS, thus a total of 52 MORE DS required just to protect the supply and rail/overland connections to Sicily.
Let that sink in – if one is serious about defending Sicily AND it’s communication links with the mainland, a conservative estimate of what that will require is 164 DS IN ADDITION to what the Axis set up on Sicily. It may be instructive at this point to note that the ENTIRE Italian army setting up in mainland Italy is only worth 129 DS; again being very generous in counting all unsupported units as supported… somehow.]