On 14 April 2018, US, French, and British military forces launched missiles against Syria, in response to a chemical weapon attack by the Asad regime against the rebel-held town of Douma a week earlier. This followed a pattern of repeated use of chemical weapons by the Syrian military, despite a previous US retaliatory attack in 2017. Three sites, all associated with the regime’s chemical weapons programme, were hit.
This led Graham Longley-Brown to ask whether the British government had conducted a wargame of the proposed attack before carrying it out:
Did anyone wargame the geo-political situation? Were expert players and decision-makers engaged to represent Syria, Russia, Iran, China, The Arab nations et al to elicit plausible reactions and risks, various categories of ‘unknowns’, and maybe even a Black Swan or two circling just out of sight… Having just re-read Tom Schelling’s Zones of Control chapter ‘Red vs Blue’, it struck me that something at the pol-mil level along the lines of his games would have delivered significant benefits to our decision-makers, both in terms of an enhanced understanding of the situation and their actual decision-making process.
If no wargaming occurred, why not? The answer to that question might shed light on the attitudes to wargaming within the MOD, despite the recent success of the second VCDS Wargame. There was time to put some combination of matrix/seminar game on the table, possibly informed by M&S and/or other wargame results. Rex Brynen suggested that even a well-facilitated and well-moderated BOGSAT would have been useful. There was time to design and play such games, either in parallel, in sequence, or both. While not quite a cycle of research, some sort of triangulation or cross-pollination would have reinforced insights arising and shaped more detailed analysis. This would all have been TBD during whatever rapid design process would have been implemented. If no gaming occurred because ‘it takes too long to develop these games’, do we need to have a bank of Schelling/matrix-like games developed for identified trouble-spots and waiting to be pulled off the shelf? This along with a wargaming ‘rapid response team’ to tweak these and then facilitate rapid gaming?
If wargaming did occur, did someone in the MOD reach for a phone and speed-dial ‘wargamers’? If ‘yes’, who was at the other end of the speed dial number? It should be (certainly include) Dstl’s Wargaming Team and Tom Mouat. Did this happen, and is the process formalised? If not, why not? Did the call recipients respond by putting a series of appropriate games on the table within hours (Mark Herman-like)? Was Dstl involved? If not, why not? Crucially, what lessons were identified with the process of rapidly designing and executing a wargame, and how will these be captured and turned into lessons learned?
Did anyone model the attack in detail? It would have taken Jeremy Smith about 2 minutes to ‘RCAT’ this using open source data, playing tunes with variables such as Russian SAMs engaging or not engaging and different permutations on Syrian AD. I suspect this would have been insightful. However, computerised sims would have been far more important. Were any used?
Finally, how might the non-MOD professional wargaming community (e.g. Cranfield) get hold of classified data from the attacks to further validate their sims? We have previously used open source data from real-world examples such as Mosul and Sirte – before, during and after those events – to validate and refine our irregular warfare RCAT models. Doing the same with near-peer, peer and peer + adversaries in a high-intensity warfighting context, will become increasingly important. What AD engaged? What was the success rate of the missiles launched? What are the BDA results? How was targeting conducted, and how effective was it? What Collateral Damage was caused? Etc. If the results are too highly classified to release then we are missing the opportunity to improve the simulations we all espouse the utility of and use for actual Defence planning. Access to real-world data such as this is crucial. Let’s hope examples remain rare, but we should leverage them when they occur.
Any such wargame would have been classified, so there’s a chance we wouldn’t know. However the consensus among UK wargamers was that no, they probably didn’t. Should they have?
One experienced UK wargamer replied:
I’ve had a think about this (with some help) and we need to be careful. There is no way we will get the great and the good to spend half a day away, in the middle of a crisis, to play a wargame.
We need them to play wargames regularly to get an appreciation of what wargames offer, but in a crisis we need a parallel process to run alongside the crisis planning. It needs someone with a trusted ear to the commander (VCDS?) who heads off and gets us lot together and then reports back with what we found…
We then need somewhere we can do that, at the right classification, that is available (like the JFC Battlelab).
As for validating our sims – I think I’m looking at a higher level (the Pol/Mil implications or the crisis and subsequent reactions, rather than the detail of the military action), so I’m less worried about building a better mousetrap, as getting insights to the commander.
The important observation here was that crisis wargaming might not (for reasons on time, among other things) be a central part of the process, but it could be a useful adjunct. To do that, however, there has to be an existing on-call capability to design, populate, run, and assess a game—quickly enough that it can raise issues for planners to consider, and with a solid enough track record that any such inputs were welcomed by planners and decision-makers. One experienced observer questioned whether the UK would ever be in a position to deploy a wargaming team quickly enough to support this sort of compressed decision-making cycle.
My own response was that this might be a case where digital simulation and modelling would be far more useful than manual games, since much hinges on the interaction of really technical variables (topography and radar shadow generated by the Anti-Lebanon mountain range, the exact placement of Syrian and Russian EW and target acquisition radars, SAM effectiveness against low flying targets amid considerable ground-clutter, and the hit (Ph) and kill (Pk) probabilities of Tomahawk/JASSM/Storm Shadow/SCAMP, and so forth).
Regarding the political dimensions of the attack, I may be a wargamer but I am not convinced this would best be explored in a wargame at all. Instead, based on my own experience, a well-moderated BOGSAT (bunch of guys/gals sitting around a table) might work better, Much would depend, of course, on the BOGAT form, the expertise at the table, and the skill of the facilitator. It is also possible that a relatively quick wargame might provide input into those discussions, something we had previously noted in our various matrix games of the counter-ISIS campaign in Iraq.
For what it is worth regarding the attack itself, I think the target set was way too narrow, and that the regime was genuinely pleasantly surprised that so little of so little value was hit: a research facility and a few bunkers. No key regime assets or capabilities were struck. None of the units or facilities involved in the Douma attack were attacked. Nothing that signals any significant cost to Asad was destroyed—indeed, the regime secured the Douma area and almost all the Damascus suburbs in the meantime. One American contributor to the discussion similarly noted “The aparent dearth of notable results makes me suspect that no one cared over much about militarily effective action but rather focused on ‘doing something’ that looked like ‘punishing Assad’ without risking a serious confrontation with [Russia]” A wargame might have brought all that out, but so too would a half hour conversation with a reasonably competent Syria analyst.
An experienced American wargamer commented that if the strike had been wargamed, the most useful insights might be in the area of coordination and process:
The real war-game here is the inter-coalition coordination of the strikes, how the C3, legalities, and permissions worked, and how the planning process leading up to the strikes would work across the different coalition partners. Who was in charge? How was the coordination done (NATO or multi-lateral)? What C3 systems were used and how did they interoperate? How did authorization flow and deconfliction occur? The technical side of the strikes is, as I believe someone said, pretty much physics and targeteering, which affects C3 but is not necessarily an interesting game in itself. I’m sure there have been many games done that look at coalition strike C3 in both a NATO and a multi-/bi-lateral context amongst the countries involved. That gaming probably informed some of the experiences and decisions of the officers making the call as to how to organized the event.
Further thoughts are welcomed in the comments section.