On Episode 910 of Effectively Wild, Ben and Sam fielded a listener question about which is the least interesting inning in baseball. The discussion naturally also turned to which was the most interesting inning, and if you could only watch one inning, which it would be; a question that’s clearly unanswerable on a number of levels.

So let’s try to answer it.

Let’s set the scene first, with a horrible dystopian vision of the future. MLB conspires to increase their full MLB.tv package and game ticket prices to exorbitant levels that only Marlins Man can afford. They instead introduce a single-inning package, which allows you to pay a much lower fee, but you must choose which inning of each game you want to be able to watch before the season starts. You can’t switch innings whenever you want, or go back and watch condensed games or highlights; it has to be only that one inning, whether on the TV or at the ballpark. What do you choose?

What would we want to see, if we had to choose just one inning in each game for the entire season? The most homers? The most runs? Action on the basepaths? Elite pitching? The highest leverage situations? Clayton Kershaw? Every walk-off? That last one might be easier to give a straight answer for, but is there any reason to think that you should watch one inning over another for those other reasons, or anything like them?

First, the Play Index Split Finder can easily show us the numbers for each inning. To start simple, here’s 2015. (Mobile users: scroll for more stats.)

Easy if you wanted to see lots of offense: just watch the first inning! Ben said on the podcast that the first featured the most runs, and in 2015 that was certainly true. The most homers, most runs, the highest OPS, comfortably the most stolen base attempts; the first inning was the inning to watch in 2015 for offensive action. As might be expected, the 8th and 9th are low-scoring, high-strikeout affairs; while you might see tense late-game action, there’s a lot less in the way of run-scoring and a much higher chance that all you’ll see is a scoreless frame, or the game meandering slowly to a conclusion with one team leading by 7. If you like to see dominant relievers strike batter after batter out, of course, then picking the eighth or ninth is excellent, but would it really still be good if that was the majority of your viewing experience?

We’ll come back to that. There’s a chance, of course, that 2015 was an outlier. Is the first inning really always one of the best for run-scoring? Again, I used the Split Finder to gather the stats for each inning, this time since 1998, in order to cover the whole 30-team era but still focus on an environment in which pitcher usage is broadly similar. In order to account for the different run-scoring environments, I plotted a graph using the tOPS+ for each inning by year, to show the relative level of production to the overall league OPS that year, therefore giving a better idea of which inning was the ‘best’ to watch for offense in each season. (Click on charts to enlarge.)

So the first inning definitely is good, but it certainly hasn’t consistently been the best. It makes sense that the first inning always has above-average batting; you know you’re getting the top three hitters in a lineup, and quite possibly the fourth and fifth hitters as well. It follows that the second is a real disappointment, consistently rating below average. The bottom half of the lineup is likely to be coming up, and it’s still going to be the pitcher’s first time through the order the majority of the time. Let’s cross that off the list if you’re looking for run-scoring. The third suffers a similar fate, albeit not to quite such an extreme extent. There’s a decent chance that one of the worst hitters will be leading off, perhaps even the pitcher, and so it can only fluctuate around average. The eighth and ninth have cemented themselves as low run-scoring environments; it’s interesting to note that the eighth has really separated itself from the second as a low-scoring inning, and even beat the ninth in 2014. Relievers are really good now.

Another interesting point to note is that the first inning tOPS+ of 112 in 2015 was the best mark here, not just for the year, but in the whole sample. Managers have started to bat their best hitters in the second spot in the lineup more often, with notable examples like Josh Donaldson and Joey Votto spending more time in the two-hole in 2015, so perhaps this is evidence that there is now a higher calibre of hitter guaranteed to come up early in the first; which, after all, is fundamentally why managers should be using their best hitters in that spot. Thus far in 2016, the first inning leads again, with a tOPS+ of 109, ahead of the fourth at 106.

Prior to this development, the 4-6 range can all make a legitimate claim, as they are consistently above-average offensive innings. Contrary to Sam’s assertion, the fourth looks like a great inning to watch for offensive action. The fourth and sixth seem to have the advantage over the fifth as the only other innings to take top spot, but in recent years the fourth seems to have slightly edged ahead; a development that I would suggest has been sparked by the increased willingness to go to the bullpen earlier in games, both as a method of reducing the number of innings and pitches that a starter throws, and to reduce the effects of the times through the order penalty, therefore making the sixth a slightly lower-scoring inning. The fifth arguably has the benefit of the lowest standard deviation in this group; you won’t see as much offense, but you’re also fairly sure of what you’re going to get.

The problem with these innings, particularly as the game goes on, is you’ll have to sit through some pitching changes. That might seem pretty irrelevant to many, but it’s probably going to seem a lot more tedious when that’s the only inning you can watch. The fourth gains another advantage in that regard, as generally speaking it will be too early to remove the starter in all but the most disastrous of efforts.

If you want baserunning action, it’s a no-contest, as the 2015 data suggests. It’s hard to quantify all the baserunning events in one stat, so I’ve used stolen base attempts as a reasonable proxy for level of aggression and action on the basepaths. Here’s a graph of the same time period but with stolen base attempts per PA (in an effort to be a little fairer to the ninth) instead of tOPS+:

Although stolen base attempts have declined somewhat, and the gap is smaller than it used to be, the first inning is still when the action happens here. Even in 2013, when the third inning was closer to the first than any other year in this sample, there were 79 more stolen base attempts in the first inning. Over the course of the season, at the rates seen in recent years, you would see around one extra steal attempt per night on average if you were to watch the first inning of every game. It’s closer to two a night if you compare it to the second or the ninth (Sam was right, the second is a bad inning).

If you’ve been reading thus far thinking that actually what you want to see is pitching dominance, or a pitching duel, it’s not hard to reverse all of this. In terms of a quality-quantity combination, the eighth is your best bet, combining a high strikeout percentage (above 22% each of the last two seasons) with the full inning that you don’t get from watching the ninth. While you often get the most dominant pitchers in the ninth – the K% has been around 24% for four consecutive years now – you also lose about 25% of the plate appearances, because many of the games are ending before the bottom half, so you’re simply losing a significant amount of baseball compared to all the other innings: about 5000 plate appearances a year. If you have a fondness for setup men over closers, like Ben and Sam, then you can mark down another point in favour of the eighth.

The second actually recovers a little by K%, rising above 21% last year, but it’s still pretty close to the fifth and doesn’t distinguish itself hugely in this regard. It certainly has the benefit of being almost as likely as the first to feature both starters, but why watch the second when you know you can see both in the first? Perhaps you like pitching changes and manager strategy: in that case you should pick the sixth, which has featured a larger number of pitchers than any other inning over our timeframe. Managers are also less likely to be rigid with their relievers earlier in the game, with those defined roles coming more into play in the later innings. They might not be as good, but you’ll see more of them (if you don’t want to see lower-quality relievers and lots of changes, don’t pick the sixth).

What are the chances that you’ll see a reasonably high leverage situation if you pick one of the last three innings? “Late & Close” are situations defined by Baseball-Reference “as plate appearances in the 7th or later with the batting team tied, ahead by one, or the tying run at least on deck”. Using the Play Index Split Finder again, I calculated the percentage of games and plate appearances that come in Late & Close situations. Since 1998, between 60% and 65% of games have featured at least one such plate appearance, but the percentage of plate appearances that actually fall into this category has only been in the 15-17% range. That means that while you’re more likely to see a Late & Close PA than not, there’s a good chance that situation won’t exist throughout innings 7-9, and it’s more likely that you’ll miss those PA with just one of those three innings in any given game.

What about specific high-leverage moments, like home runs? The Event Finder displays events such as home runs by high leverage, which is defined by Baseball-Reference as any situation in which the Leverage Index is over 1.5, so I ran the finder for every high leverage home run in our sample period, then calculated the percentage of the total for each inning:

So I guess we really shouldn’t watch the second inning. As you’d expect, the chance that one will be hit goes up as the game goes on, but innings six through nine are pretty close in terms of the likelihood that a high-leverage bomb will be hit. 1,117 of those homers were of the walk-off variety, out of 15,312 total home runs. That accounts for almost all of the walk-off homers in that period, but 81 were categorised as medium-leverage (0.7-1.5 LI) so in total, you’d have missed 1198 walk-off homers if you’d never watched the ninth inning since 1998.

The play on the field isn’t the only factor we need to think about, though. BttP’s Brandon Lee suggested thinking about crucial broadcast and ballpark factors. Perhaps your team has guest announcers in certain innings, or special guests singing in the seventh inning stretch. These may be crucial factors in you deciding to watch, or avoid those particular innings. Do you love the giant-headed mascot or sausage race? What about getting free pizza when the pitcher strikes out the side? Does it get too cold late in the evening at certain parks, or is it better to go late because it’s usually too hot when the game starts? Is the traffic a disaster if you go for the start of the game, and non-existent when you turn up in the eighth? If you were going to the park but had selected a later inning, you would presumably need to hover somewhere: perhaps, in keeping with our dystopian theme, some kind of series of TV-free holding pens, or risk missing your inning.

Ultimately, you’d have to decide what you value most. As Sam pointed out on the episode, the outcome is often decided by the ninth, but Ben said he’d rather have the ninth, simply because you always get the outcome of the game, and sometimes that will come in thrilling fashion. If you get the bulk of your enjoyment in baseball from watching particular players, that would probably dictate your choice the most. If it’s a starting pitcher or great hitter, it almost has to be the early innings, and the first is the only inning where you have a guarantee of knowing what the starting situation is going to be. The further you get into the game, the lower the chance is that you’ll actually get to see a particular starting player. If you spend all your time watching replays of relievers throwing ridiculous strikeout pitches, choosing the seventh, eighth or ninth is much more logical, as long as you’re prepared for a lot of those innings to be meaningless in the context of the game result and therefore perhaps being pitched by the sixth or seventh guy in the bullpen, not the relief aces.

There’s also something unsettling about the idea of selecting the seventh or eighth for the prospect of seeing a tense late situation unfold, only to be unable to watch the game’s conclusion. Imagine that every time you watched a game going into the ninth inning with your team tied, you had to turn it off and not watch the resolution. Imagine if someone was pitching a no-hitter or perfect game, and you stopped watching when they were 3 or 6 outs from finishing it off. Imagine knowing that you missed every single walk-off, every game-ending bases-loaded strikeout, every game finished without a save, by just a few minutes. Would it be worth it? Sometimes the game would be all but over after you’d watched your inning, but there might be a 5%, or 2%, or 0.1% chance the trailing team could come back, and you’d always miss it when it happened, and you’d always feel like it was tantalisingly close.

For me, it’s all or nothing. If you love just watching baseball, taking in the action and the atmosphere, it has to be the first. You get the most offensive production, the guarantee of seeing both starting pitchers, virtually no chance of sitting through pitching changes, and none of the disappointment at having to stop watching a game immediately before its conclusion. If the result is all-important to you, you have to watch the ninth. It doesn’t matter that teams trailing after eight win just under 5% of the time, as BttP’s Rob Mains noted earlier this year. You don’t know that the team will win, and you won’t get the euphoria, satisfaction or relief of watching a comeback unfold or a reliever shut down a rally under the highest pressure to win a game. Maybe that doesn’t matter after a while; maybe it stops making a difference to you if you never get to watch a game end as long as you know how it ended. Much of the time, you will see the most crucial moment of the game in the earlier innings, but does that mean anything if you aren’t watching the whole game?

In the end, watching an inning might never be enough. How much is lost when the context of the other eight innings is stripped away? The effect might not be that you stop caring that you never see the end of the game, but that, divorced from the majority of every game on a daily basis, you stop caring altogether, and stop watching altogether.

I won’t presume to know how other people watch baseball, so instead I’ll make some recommendations based on a few different criteria, depending on your viewing habits and personal philosophy, just in case you ever find yourself in a baseball-centred dystopia.