Both Jeff and Randy, in part, advocate a proactive Draft strategy, be it because of a personal colour preference or, in some cases, due to an inept drafter on your right. They do this only to include every method of drafting. For me, in accord with Randy, the only time I let colour preference influence my decisions is when I am faced with a decision between two cards of equal value early on in the draft when I have no other information to go on. The deeper you get into the draft and with every pick you get passed, you are fed more and more information as to what colours the people on your right are drafting and, more importantly, what colours you should be drafting. In short, the only way to draft correctly is to do so reactively.
There are a few important things to take into account concerning your first pick, before the reactive nature of a draft kicks in. The first thing to think about is that your first pick means nothing. It should have almost no influence upon your draft. Sure, your next few picks will be slightly biased towards trying to keep your first pick playable, but if it isn't in the colours you are being passed, then it wasn't meant to be. The same almost goes for your second and possibly third pick because when you make them you still have little information to go on.
Always maximise the potential of your first pick. This means taking the most powerful card. Early on you do not have the context to place a card's strengths upon the background of other cards you've drafted to see if it works or fits a theme, you just have its raw power to account for. If two cards are very close in power level, then two factors can influence your decisions—colour preferences and likelihood the card will make your final deck. Colour preference is simple: some colours might be stronger than others, or suit a player's style more, enough to warrant a slight priority shift. Some cards are just more likely to make your final deck than others, cards that are inherently splashable, are artifacts, or have a low colour commitment.
If you are faced by a weak pack with the option being between Prismatic Lens and Amrou Seekers, you pick the lens because it will almost always make you deck. Stronghold Overseer versus Lightning Axe is an example of a pick where one card is simply too powerful to pick a more easily playable card over it. Mindless Automaton versus Rift Bolt is a closer example. They are very close in power; for example, I prefer the Automaton whilst Julien Nuijten opts for the Bolt. Both cards have a high likelihood of making the cut at the end of the day—the Automaton is slightly more likely, but the Bolt is easily splashed. Here, even though it is a very close call, I think the Automaton is the pick as it is will always be played, whilst the Bolt still has a chance that it might not be. Might Sliver against Knight of the Holy Nimbus is another classic example. The Knight represents everything that is colour commitment and is a better card than the Sliver, whereas the fattie is a slimmer commitment whilst opening up the doors to the potential of the Sliver archetype. Colour preference can also factor in here as white is somewhat weaker overall than green in this format. It is for the potential that I advocate the Sliver, but it makes the cut because it is more likely to be played.
There is another mistake that I see players making a lot. Say your first pick gives you the options of Errant Ephemeron, Fathom Seer, Looter il-Kor, or Lightning Axe. Many people advocate taking the Axe and staying away from the ensuing blue bloodbath, even though the Ephemeron is the slightly better card. In my opinion, they are wrong, for various reasons. Firstly, the Ephemeron is the better card and you always want to take power early on. Secondly, if you take the common dragon, then the guy on your left will be faced with the two card drawers and the removal, and will correctly take the Axe, so the expected grabbing of blue cards won't happen until further down your sphere of influence. Thirdly, and this is the biggest reason coupled with it being a better card, you shouldn't care what the guy on your left is doing! If you can help it, then be all nice and dainty, but never do so as a sacrifice to you. He'll only be passing one pack to you, and even if he did pick the Seer to follow up an initial-pick blue card, and will hence be cutting you in the second pack, it might not even be of consequence to you. You might even have abandoned the Ephemeron when blue fails to come from the right. At the end of the day, the guy on your right is your master of your fate whilst the guy on your left is none of your concern.
The success of your draft is entirely dependent on how well you read the signals coming from the right. The only factors under your control during a draft are the cards you pick and, as a consequence, the cards you pass to your left. If you ship certain cards in a particular colour, then it is safe to assume that the person on your left will be in that colour. However, if one of the colours you were initially taking dries up and is instead replaced by the colour you have been passing, you should switch into it. Many players are hesitant to switch into this colour knowing that you will see less cards of quality in the second pack. However, what the player on your left does is of no import to you. Sure, if you are faced with a colour choice and you can reliably assume that one of them is not being drafted by the player downstream, then take the other colour if all other factors are equal. It will almost never be the case, from the third pick or so on, that there will not be other, more influencing, factors. I, of course, speak of the signals you're receiving.
The only signal you have for your second pick is to work out what card is missing. In past sets this has been fairly easy, as there was only one rare, three uncommons and eleven commons. Now we have a "timeshifted" card in Time Spiral and foil rares taking the place of a common slot. However, these shouldn't be enough to put you off which card the guy on your right picked. You also know that, assuming the guy is a good drafter, that he picked a better card than any other card in the pack. If he picked a common and there is still a Lightning Axe in the pack, it's pretty obvious he picked an Errant Ephemeron (the only common you would pick over the Axe). You will not often be able to deduce the exact card that was picked, but you can often narrow it down. For example, if there is still an Ephemeron in the pack and, say, a Greater Gargadon, and there is an uncommon missing, there are only really three cards that could have realistically been taken—Tromp the Domains, Firemaw Kavu, and Sulfurous Blast. Two of the three are red, so until showed otherwise, it is a relatively safe assumption that he picked a red card.
Unfortunately, even if you can deduce the exact card picked, the hard work is far from over. Just as you must adapt to everything the player on your right does, so must he in turn react to everything the guy in his right does. This means that as your first few picks are uncertain, so are the guys on your right. If you work out that he picked an Errant Ephemeron, and you are then faced with the option of Looter il-Vec or Durkwood Baloth, it might still be correct to take the Looter if your first pick was Blue. As with your first pick, tuned slightly with every bit of new information you receive, you should still take the most powerful card in each pack until you are sure what colours are open. This might often result in your first four picks all being different colours. This is a good thing—sure, you would rather they were all in the same colour, but you now have the added flexibility of still being adaptable, whereas many worse drafters would've dogmatically stuck to their first pick and would have doomed their draft before they even cracked their second pack.
One last important tactic that will enable you to maximize your flexibility and hence maximize your chances of picking up on the correct signals is to make passive picks at the correct time. I've advocated that your early picks should consider power over playability, and that when the power level is close to be as passive as you can. This is especially true when you are faced with mediocre cards, including fixers. The fixers will probably make your deck, and even if they don't, for the rest of the draft they will leave open the option to splash powerful cards or help ease an otherwise shaky mana base.
I've said it before and I'll say it again, you should remember every card you pass. Not just the cards, but the composition of each pack—not because it means you can figure out what the player on your left will take each time or because you are remembering the tricks you pass to play around later (although these are valid reasons) but so you know which cards you should pick later on. Each pack you get passed is a piece of the puzzle that you need to solve to decipher which colours you should be. Even six picks in, you should be able to recall the exact makeup of every pack, so that you can reflect back upon them and see if you might have missed something or to verify that, given some irregular packs, a colour may have been open even though it didn't really appear to be at the time.
This filtering of data should also be done during the minute-long review period to make sure you didn't miss anything. It should also be noted that whilst you review, you may often be looking at a veritable pile. You might have abandoned most of you first picks and switched colours late in, meaning that you only have six or seven playables in the colours that you will end up playing. You need not worry. The beauty of the submissive strategy is that you will now be set up for the entire draft and will reap the rewards in the third booster.
Sometimes you will be lucky enough that your first few picks are also in the colours you should be drafting. When this occurs you are given the golden opportunity to exclusively cut off a colour from the players on your left. This will sometimes mean taking a slightly (only slightly) inferior card in the cutting colour over its companion shade, but it will have a two-fold effect. Firstly, it will ensure that you will get a strong flow of cards in the colour from the left next pack, meaning your deck should be awesome as your dominant colour will be flowing from both directions. Secondly, it will make that colour the dominant of the two, which will help your mana base, make certain close picks later on easier, and aid mana-intensive cards such as Shade of Trokair. If your draft was off to a rocky start so that you only settle into your colours late on, then there will still be a cutting effect, albeit a much weaker one.
When the first pack is finished with you must acknowledge that the signals you will receive from the left will only reflect what you passed him. You should avoid dipping into the other colours that now flow towards you. If you failed to cut they will probably not be exclusively your colours, because it will not help you when the third pack comes around. Not only will you be in the wrong colours, but your colour switch will have rocked the boat and you may have ruined any hook-up you spent the first pack establishing. Assuming your read on the draft is correct, stick true to your decision and ship all the off-colour goodies to your right to further lock him into his colours so that pack 3 will be the gold mine you want it to be.
What I have said here holds true for all draft formats, but sometimes there have been other factors to consider. In the full Odyssey block, Torment was heavily biased towards black, meaning that cutting black in the first pack or dipping into if it flowed in the second were perfectly valid options. However, this is rarely the case and even when it is, it is still often incorrect to do so. R&D are a clever bunch of people, and it is unlikely that they make the colours so unbalanced that sheer colour preference is stronger than the perfect colour hook-up. Hence, a mainly dominant strategy is doomed to failure as it ignores the information you receive and focuses on hoping the draft will work out for you. The reactive strategy is more successful because it takes advantage of every part of the draft you have a chance to control and makes use of all the information at your grasp.
Read the signals,