Thread by @FOSSpatents

#Microsoft and the @CMAgovUK will clash in the @CATribunal on Tuesday, May 30, for the first time: for a case management conference.

The CAT has just published a summary of Microsoft's grounds of appeal. Note that #ActivisionBlizzard also appealed, just not published yet.

🧵1/X

The court's summary is here:
catribunal.org.uk/sites/cat/file…

I will now discuss the grounds of appeal. Most of them are exactly as others and I had also expected. Some are interesting and were not foreseen in that particular form.

🧵2/X

Ground 1 attacks the part about Microsoft's "current position in cloud gaming services", starting with market definition, which is almost always the "holy grail" of antitrust cases. Think of Epic Games v. Apple (all mobile game distribution v. iOS-only app distribution).

🧵3/X

3a is an argument about "potential switching to native gaming [i.e., locally installed games]". Substitutability means products/services fall into the same market. If a user didn't get a particular game on xCloud, but could get it locally, then the market is broader.

🧵4/X

The broader the market, the lower a given player's market share typically is.

3b also attacks the market definition with a view specifically to vertical foreclosure (the CMA's concern), arguing that there are "out-of-market constraints from native gaming."

🧵5/X

To be precise, 3b is a fallback theory if 3a failed. If 3a succeeds, the market is defined more broadly (meaning an overall games market as opposed to just a cloud gaming market) and the CMA lost that part. 3b is an alternative if the CMA's cloud gaming market stands.

🧵6/X

So the 3b fallback says: even if the market is just cloud gaming, there are out-of-market [competitive] constraints. What does that mean? Microsoft's decisions (such as where to offer ABK games) would be disciplined by the fact that gamers might just play locally.

🧵7/X twitter.com/i/web/status/1…

If the CAT found that the CMA had to consider those out-of-market constraints (market = cloud, out-of-market = locally installed), the vertical foreclosure theory (withholding games from other cloud services) would have to be redeveloped.

🧵8/X

Sorry, I just realized a typo in the previous tweets: I meant 1a and 1b, not 3a and 3b. I'll leave those tweets as many people have already liked them etc. and the specific meaning is discussed, so no confusion there.

🧵9/X

The final part of Ground 1 is 1c. It argues "fundamental errors in its calculation and assessment of market share data". Presumably that involves things such as counting all GamePass Ultimat users as xCloud users, but not all Prime users as Luna users or so.

🧵10/X

What if Microsoft wins any part(s) of Ground 1? That alone would presumably lead to a remand and the CMA would have to make a new argument based on Microsoft having a much lower market share and/or its conduct being disciplined by out-of-market factors.

🧵11/X

The CMA could try to block the deal again, but it would then be at a heightened risk of the next decision again being irrational. It couldn't play the untruthful PR game of saying "Microsoft already owns 70% of the market". All sorts of incompetent people echoed that one!

🧵12/X twitter.com/i/web/status/1…

But Ground 1 could also succeed in combination with other parts. Then there could be a situation where the CMA can't block again without making itself so obviously ridiculous that it can't even spin its story for PR purposes.

🧵13/X

In Epic v. Apple, where a petition for rehearing with a U.S. appeals court is due in about 2 weeks, I'm very much in favor of focusing on market definition. In this CAT appeal, I like the market definition argument, but it may not kill the CMA's case all by itself.

🧵14/X

Ground 2: CMA didn't "take proper account f three long-term commercial agreements ... with cloud gaming providers". I guess Nvidia, Boosteroid, Ubitus (the first three). The summary doesn't explain WHY Microsoft argues the CMA didn't properly consider those contracts.

🧵15/X

Having read the CMA decision, I felt that the parts where they dismissed those agreements as reliable and useful showed the CMA's bias very clearly. That bias shines through most pages of the decision but how they dismiss the significance of those contracts is outrageous.

🧵16/X

What if Ground 2 succeeded alone? CMA could still find excuses for blocking, but it would be harder.
If Grounds 1 & 2 succeeded together, the CMA's ability to block again would already be severely limited. Really severely.

🧵17/X

Ground 3: irrationality and procedural unfairness concerning the holding that Activision would've been likely to make its games available on cloud gaming services.

Yes, irrational to rely on non-decision-maker employees. Also, simply no real evidence.

🧵18/X

If Ground 3 succeeded on substantive grounds, it would mean that the merger is actually good for cloud gaming as a whole.

If only on procedural grounds, CMA could try to arrive at same conclusion again.

🧵19/X

Let me stress here what I've said on various occasions: it's an "urban legend" that the CMA can always just arrive at the same results again. Under the LAW (Enterprise Act 2002) it MUST decide in accordance with the CAT's guidance. It's written in the law!

🧵20/X

What about permutations of wins involving Grounds 1-3:
If all 3 of them succeeded, CMA could theoretically try to block again, but it's hard to see how it would work. If I had to pick 1, I'd say Ground 1 though in a merger case not as powerful as in past-conduct cases.

🧵21/X

Ground 4: foreclosure analysis (withholding games from rival services) is attacked as being unlawful. Four errors are being challenged, arguing that any one of them, but least some or all of them would defeat the CMA's foreclosure analysis.

🧵22/X

a) "wrongly relied on evidence that so-called 'AAA' games would be important" to make that finding specifically for Activision games. In other words, even if AAA titles like FIFA, GTA are generally important, that doesn't mean CoD or other ABK titles necessarily are.

🧵23/X

b) CMA didn't take account of relevant evidence of immediate losses from such foreclosure: in other words, like with the console gaming theory, the CMA (with its obvious bias!) focused only on benefits and not enough on expected costs of anticompetitive behavior.

🧵24/X

c) failure to take account of (i) out-of-market constraints (impact of native games; wouldn't play a role if market is already defined broadly anyways under Ground 1a) and (ii) the cloud-gaming agreements already concluded.

🧵25/X

If Ground 4 succeeded, the CMA would basically lose its cloud gaming theory of harm the way it lost the console theory: its theory that it would be bottom-line positive for Microsoft to engage in foreclosure would be killed. It could, however, somehow try to rebuild it.

🧵26/X

Ground 5: this is in my view the best shot at defeating the CMA for good. I've said before--here on Twitter--that if the CAT found that the CMA was wrong to reject the proposed remedies, it's game over for the CMA even if the case goes back to CMA in a formal sense.

🧵27/X

I most recently reiterated that position yesterday in my podcast with @MrboomstickXL @EverbornSaga @PostUp_SOG. The other grounds of appeal are all very good, but this is the one that more clearly than any other can checkmate the CMA on a definitive basis.

🧵28/X

They again have multiple attack vectors for a single ground of appeal: 5a-5d.

5a: legal error in the form of believing that a "comprehensive remedy" (block or divestiture) was the CMA's duty.

🧵29/X

5b: failure to consider comity. That word means the international, diplomatic, and trade-related aspects. Comity--or comity of nations--means countries respecting each other. Here, the CMA is crazy and wants to block a global deal instead of imposing UK-specific remedies.

🧵30/X

Not only the CMA's absurd decision but also its CEO's public statements after the decision show that the CMA was hell-bent to just play the world's policeman for mergers, particularly for Big Tech mergers.

CAT may tell the CMA that such megalomania is not rational.

🧵31/X

5c: specifically about CMA's rejection of Microsoft's remedy (the one the EU accepted, though that does not mean binding precedent) as "disproportionate". Very important word! Why? Because of the standard of review:

🧵32/X

Microsoft's grounds of appeal generally seek to avoid stressing the word "irrational" too much, given that it's a bit hard for an appeals court to slap the agency, saying it acted irrationally. Disproportionality has same effect (Microsoft wins) but sounds gentler.

🧵33/X

Especially common law (UK, US) judges such as the ones serving on the CAT are very much focused on proportionate, or also equitable, remedies. So this appeals to the judges' professional honor to ensure that the impact of a ruling should not be disproportionate.

🧵34/X

5d: Again, common law: "breach of [CMA]'s common law duty of fairness and the CMA's own remedies guidance".

It's true any of 5a-5d could solve the problem. 5c is the one I like best. 5a/5b/5d could pave the way toward prevailing on 5c. If 5c works, the CMA is CHECKMATE.

🧵35/X

The court's summary mentions Weil, Gotshal & Manges, the firm that also made Microsoft's filings with the CMA, as its counsel. But as we all know, Daniel Beard KC of Monckton Chambers is on the case and will then appear at the hearing, I guess.

🧵36/X

The summary suggests that Microsoft is being diplomatic, trying to avoid "CMA-bashing", though they'll obviously have to--and presumably have--made lots of arguments that show the CMA was biased, unfair, and didn't understand the market or the technology at all.

🧵37/X

I believe the two grounds that "bookend" this appeal -- Ground 1 on market definition, market share, competitive constraints and Ground 5 on remedy -- are the most important ones, but the parts in between the "poles" can also help in various ways.

🧵38/X

The case management conference next week (presumably livestreamed) will be interesting and the question is then going to be when the most important hearing -- the one after which the court decides -- will be held.

🧵39/X

As I predicted, Mr Justice Marcus Smith, the president of the @CATribunal and also a #patent judge, will be presiding over the proceedings.

For now he's the only judge listed, but there'll be a panel. Two others will join him.

This appeal ROCKS

fosspatents.com/2023/05/antitr…

🧵40/40