[Free Podcast] So Many Insane Plays # 43: Gitaxian Probe and Interview with RayR

[evouga]

Excellent as always.

I loved the discussion about how more information can sometimes be harmful, not because mathematically more information is ever harmful, but because our brains are not equipped to make correct plays with this information. I'm reminded of the common quip that Doomsday is the best deck in Vintage, but doesn't perform as such at tournaments because nobody can play it perfectly.

But setting aside the perfect-play objection, I'd like to explore a bit the claim that more information is better. Certainly, if appropriately formalized, this is true: given perfect play, gaining more information can only increase your probability of winning the game.

But Probe doesn't just give you information. It gives your opponent information as well. Some of this information is pedestrian -- you are playing a deck with Probe, and had Probe in your hand. But it also gives your opponent the knowledge that *you know your opponent's hand.* This is an incredibly subtle point and best illustrated with an example.

Both players have plenty of mana and the following cards in hand:
Player A: Gitaxian Probe, Phyrexian Revoker, 3 blanks
Player B: Time Vault, Voltaic Key, Force of Will, blue card.
It is Player A's turn, and both players know the opponent is playing a blue-based deck. Consider these scenarios:

Scenario 1: Player A plays Probe. The game is now forced, with Player B winning: Player A must cast Revoker, and Player B, knowing that Player A will name Vault (or Key), must counter it, and then can play Vault+Key unmolested.

Scenario 2: Player A casts Revoker blind. Player B may counter, or may not, reasoning as follows: Player A has a number of cards she might name with Revoker, such as Black Lotus, Dack Fayden, or Jace. With some probability it is correct to let Revoker resolve in order to fight a potential counterspell when playing Vault+Key, and take my chances that Revoker won't shut off the combo.

So even with perfect play, Player A is worse off after playing Probe. Note that the key element here is that Probe allows information to flow in both directions. Contrasts to a final scenario:

Scenario 3: Player B is wearing mirrored shades, and Player A gets a surreptitious peek at Player B's hand. Player A now runs out Revoker. If Player B doesn't counter (and it is correct for him not to, with some probability), Player A shuts off Vault+Key and is in the best position relative to the other two scenarios.

[Chubby Rain]

snip

So your argument is that using Gitaxian Probe enhances the value of Phyrexian Revoker, but your opponent knows this and is more likely to counter it. The counterpoint to that is increasing the value of your own cards is generally not a bad thing and if you had a single piece of additional interaction with your opponent (flusterstorm, steel sabotage, misstep), the above situation works out much better for you and much worse for your opponent. We can always create hypothetical situations but the important thing is how likely these situations are (for example, the trap of the "best-case scenario"). While it is interesting to take note that the opponent does know what you know about their hand, it seems that most of the time the information will be to your benefit.

Personally, I like Gitaxian Probe as most counterspells are situational now (Flusterstorm, Pyroblast, Mental Misstep, etc.) and having more information about your opponent's lines helps you to optimize your own (what cards to scry with preordain, what cards to discard with Dack). That probe is also a sick follow up to Mentor and fuels Delve justifies its use as a 1-3 of in these decks, in my opinion.

[evouga]

That's right. My point is that even if Gitaxian Probe were absolutely free, it gives information to both players and so it is not always true that playing it logically always increases your probability of winning, even given perfect play.

But that doesn't mean it isn't more beneficial than harmful, most of the time. I leave that judgment to the Vintage experts.

[JarofFortune]

 I'm reminded of the common quip that Doomsday is the best deck in Vintage, but doesn't perform as such at tournaments because nobody can play it perfectly.

I find this very hard to believe considering the deck's abysmal shop matchup.


The podcast was excellent. The probe discussion was a great example of the type of content we need more of. The interview was incredibly eye opening. Whenever I have heard someone mention a TMD open in the past(a rare occurrence) they have done so as if the events had a mythical quality. Now I know why.

I think Sensei's top would be a good card to cover sometime. In the same note, I like the idea of a cast on managing your manabase in tandem with cantrip sequencing(Holding or cracking fetchlands before cantripping; planning ahead a turn to fetch the right color, and how your actions change depending on what cantrip you cast). I'd be very interested to hear your views on both of these, especially the latter.

[Chubby Rain]

That's right. My point is that even if Gitaxian Probe were absolutely free, it gives information to both players and so it is not always true that playing it logically always increases your probability of winning, even given perfect play.

But that doesn't mean it isn't more beneficial than harmful, most of the time. I leave that judgment to the Vintage experts.

Or Legacy experts, where Probe is used somewhat ubiquitously. The card has a proven track record in every other format in which it is/was legal.

[Protoaddict]

On the topic of cards to discuss, the card that has given me the most trouble in recent years has honestly been sol ring, and I know that sounds like crazy sauce but besides shops that will always use it and things like dredge that will never, the card has so much grey area and so many facets that I always wonder if a deck should run it.

It is good against shops just to pay taxes or spell pierce mana
Sometimes it is better than off color moxen, sometimes not depending on your decks make up
It can open you up to misstep or just be another target that you use to draw out a misstep
Depending on how much mana you need off color and how many spells you use a turn it can be back and forth on if you need to include it or not.
It can be a sideboard card against shops or mono white stax decks if you think about it, and can be good in sandbag sideboards.

I think that would be a good discussion.

[Will]

I really enjoyed this podcast.

The idea of what is being cut for Gitaxian Probe is quite interesting. While discussing his card choices in Delver, Craig Berry and I went back and forth on Gitaxian Probe and had some slightly different points of analysis than you mentioned.

One of the things that determined the inclusion or lack of Gitaxian Probe for Craig and I was Snapcaster Mage. Being able to play Snapcaster as a 2 mana 2/1 which peeks when it comes into play is quite good. With Snapcaster present in the deck, Craig felt that Gitaxian Probe was worthwhile as at least a 1 of before the tournament.

How strongly you trust your abilities to read your opponents was a big consideration. On the drive home from the tournament, Craig said that he would have not played the 1 Gitaxian Probe that he ran because he always knew what his opponent had in their hand by reading them. While this is undoubtedly at least somewhat of a hyperbole, Craig is an incredibly skilled player and has played A LOT of Magic, so I trust his gut over most others in the community. On MODO, Gitaxian Probe is much better because it is significantly harder to read your opponent over the computer outside of measuring how long it took them to respond. For Craig and I, this was a very important aspect which tipped Gitaxian Probe from playable to not for him.

I think that the point which was made about what would be cut for Gitaxian Probe is interesting. I thought that Probe could be played instead of a 4th Dig Through Time, but this is very situation dependent. Subconsciously, the idea of a lack of space played a big part in our decision making process as we eventually dropped from 3 to 1 and ultimately 0 Gitaxian Probes in Craig's Delver list.

[evouga]

I view the "lack of space" objection with a healthy dose of skepticism. Like CDawg argued above, there is nothing magical about the number 60: if the DCI brought back 40 cards as the minimum for constructed formats, would you continue to play with 60-card Delver decks? Or would you somehow find the 20 (!!) cards to cut? I rather strongly suspect the latter. Even for decks with relatively homogeneous power level distribution, less cards means less variance and more consistency.

Similarly, *no* competitive deck plays 64 cards "because there is no space to cut anything."

Now I appreciate that one cannot simply build a 56-card deck, then throw in 4x Probe, because of the increased difficulty of mulligan decisions, excessive life loss, decreased effectiveness of card filtering, vulnerability to Spheres, Chalice, and Misstep, etc etc. But simply throwing up one's hands and declaring "there is nothing that can be cut" is not, in my mind, a compelling argument.

[diopter]

Effectiveness of the mulligan is one of the most critical aspects of a competitive deck, and not something to gloss over or "yeah but". It's the crux of the issue with Probe.

This goes double for the tempo-control strategies that must be able to ship a superficially passable hand that actually does nothing, for a hand with action and a better curve.

It goes past the numbers - one must acknowledge that even if one has a slight probabilistic edge with quad laser Probe, one will probably piss it away with one wrong decision on a no-land Probe hand.

Anyways, for me it begins and ends with the mulligan. In my experience, an easy way to lose more is to construct a 56 card deck, add 4 Probes, and expect it to work.

[Stormanimagus]

Effectiveness of the mulligan is one of the most critical aspects of a competitive deck, and not something to gloss over or "yeah but". It's the crux of the issue with Probe.

This goes double for the tempo-control strategies that must be able to ship a superficially passable hand that actually does nothing, for a hand with action and a better curve.

It goes past the numbers - one must acknowledge that even if one has a slight probabilistic edge with quad laser Probe, one will probably piss it away with one wrong decision on a no-land Probe hand.

Anyways, for me it begins and ends with the mulligan. In my experience, an easy way to lose more is to construct a 56 card deck, add 4 Probes, and expect it to work.

100% agreed. Mulls are the reason probe does NOT = 56 card deck. If probe read as it does now but had the clause:

"If this is in your opening hand you may look at the top X cards of your library where X is the number of probes you have in your hand before declaring mulligans"

or something more concise along those lines, then we would be talking about a card that is strictly best as a 4-of in most blue decks. There would be almost no drawback at all.

But this isn't the way the card is worded so it's not the way it works. Hence this isn't an auto-4- of at all.

I think it is a 4-of in some TPS builds simply because those decks value the info enough and run many top deck tutors that combo very nicely with probe. I don't see it as an obvious 4-of almost anywhere else (Well, Belcher sure).

-Storm

[evouga]

This still would not address Steve's point in the podcast, that Probes dilute the power of Preordain et al.

A true auto-4of would exile itself from the deck when you begin the game, a la ante cards.

[CHA1N5]

I'm really loving this discussion.  It speaks to the complexity of the issue that many people have a "most important" aspect of the card that they focus on.

[Islandswamp]

This still would not address Steve's point in the podcast, that Probes dilute the power of Preordain et al.

A true auto-4of would exile itself from the deck when you begin the game, a la ante cards.

I wrote some stuff about the card, and that was something I mentioned. If the card somehow let you start the game with only 56 cards in your deck, then things would be different.

[rcwraspy]

Great podcast! It was very interesting to hear such an in-depth analysis supported by game theory on a single card.  Of course the discussion has many branches into other topics as well, which is part of what makes it so interesting.

At the end of the podcast you asked what card or cards we would like discussed should you continue this type of analysis.  How about a broad analysis of the various counterspells available in the format and their applications?  Drain, Force, Daze, Fluster, Pierce, Snare, etc. are all very different and only marginally fungible.  Some decks can utilize each, others can't.  When and where do they become obvious inclusions vs. potential inclusions vs. left out?  

I know this is a broader topic than just a single card focus like Probe but I think it could be a manageable conversation if organized correctly.  The answers may also seem obvious to many of us, but so did Probe's applications before hearing this podcast.  Yet it was still an interesting and informative discussion.

[Killane]

I second the idea of a podcast is this same vein regarding countermagic in modern vintage, espeically given Steve's proclivity towards the use of spells like Misdirection in the VSL i feel like you guys would have some really interesting insights.

[Naixin]

After testing more with Probe, I have come to the conclusion that for Gush type decks (mentor control or delver), I'd rather play 2-3 probes when playing online and 0-1 probes when playing in person. I feel like yes, it does dilute the power of the deck but the information gained is extremely valuable and more than makes up for the dilution IMO. You can't read people online, hence more probes but you can in person.

[Chubby Rain]

After testing more with Probe, I have come to the conclusion that for Gush type decks (mentor control or delver), I'd rather play 2-3 probes when playing online and 0-1 probes when playing in person. I feel like yes, it does dilute the power of the deck but the information gained is extremely valuable and more than makes up for the dilution IMO. You can't read people online, hence more probes but you can in person.

The information that you gain is valuable and all but as a zero-cost cantrip, the card is absurd with Monastery Mentor. Bish Coddler's 4-0 and 3-1 decks from this past weekend have featured the full set of Probes and Mentors http://magic.wizards.com/en/articles/archive/mtgo-standings/vintage-daily-2015-07-06.