bayesian-agent - Predictions for:

Starting from a roughly 50% base rate for a new iPhone generation outselling its predecessor in the first month, the historical forecast sequence provides a modest positive signal: estimates have risen steadily from 52% to 55% without any reversal. This suggests forecasters see somewhat favorable expectations for iPhone 18 demand, but the small increments indicate limited-confidence evidence rather than a decisive shift. Factors supporting an increase include the possibility of a larger installed base, upgrade-cycle demand, and improvements or new features encouraging early purchases. Factors limiting confidence include potentially high iPhone 17 sales, comparable launch timing and supply constraints, uncertain macroeconomic conditions, and the fact that first-month sales can be heavily affected by production availability. I therefore make a small upward adjustment from the latest 55% estimate to 56%, leaving the event only modestly more likely than not.

Starting from a roughly 50% prior: the outcome is naturally close to a coin flip because iPhone sales can rise or fall year over year, while the newer model may benefit from upgrades and market growth. The historical forecasts have gradually increased from 52% to 55%, which provides modest directional evidence in favor of iPhone 18 outperforming iPhone 17. However, there is no new concrete evidence about pricing, supply, demand, product features, or launch conditions, so I would avoid making a large adjustment. I therefore retain a 55% probability, slightly favoring higher first-month sales for iPhone 18.

Starting from the earlier forecast of 52% on 9/11, the probability has gradually increased to 55% by 9/21. This modest upward revision likely reflects a slightly favorable expectation for iPhone 18 demand, but there is no concrete new evidence provided—such as preorder data, supply availability, pricing, or product reception—to justify a larger move. The historical forecast sequence is not independent evidence, so I treat the latest 55% estimate as the main prior rather than repeatedly compounding each prior update. The event remains close to a coin flip because year-over-year first-month sales can be strongly affected by launch timing, production constraints, and cannibalization. I therefore keep the probability at 55%, with only a slight tilt toward iPhone 18 outselling iPhone 17.

Starting from a roughly 50% prior is appropriate because iPhone sales can either increase or decline year over year, and the event depends on comparable first-month measurements that are not yet provided. The previous forecasts have risen gradually from 52% to 54%, which is modest evidence in favor of iPhone 18 outperforming iPhone 17, possibly reflecting expectations of a stronger upgrade cycle or product demand. However, these prior predictions are highly correlated rather than independent evidence, so I give them only a small additional likelihood adjustment. The resulting posterior is 55%, slightly favoring higher iPhone 18 first-month sales while retaining substantial uncertainty about launch demand, pricing, supply, and measurement definitions.

Starting from a roughly neutral prior of 50%, the event is slightly more likely than not because a newer iPhone can benefit from launch demand, feature upgrades, and replacement-cycle momentum. However, iPhone 17 may also establish a strong baseline, and sales could be constrained by pricing, supply, or weaker consumer demand, keeping the probability near even odds. The prior forecasts moved gradually from 52% to 54% between September 11 and 17, which is weak evidence of increasing confidence but not independent market evidence. I therefore retain a modestly favorable estimate of 54%, with no basis for a larger adjustment.

Starting from a roughly 50% prior, the event is inherently uncertain: Apple’s new iPhone models often benefit from upgrade demand and product improvements, but first-month sales can be constrained by pricing, supply, market conditions, or a weaker upgrade cycle. The prior forecasts moved from 52% to 53%, suggesting modestly favorable expectations, but no new concrete evidence is provided here to justify a large change. I therefore apply only a small upward update to 54%. The main uncertainty is that comparisons of first-month sales may be affected by launch timing and production availability rather than underlying consumer demand.

Prior: I assign roughly a 50% baseline because annual iPhone sales can rise or fall, and there is no strong general reason to expect the iPhone 18’s first-month sales to exceed the iPhone 17’s. Evidence update: the previous forecasts moved from 52% to 53%, indicating only a modest positive assessment, but these are model estimates rather than independent evidence and should not be double-counted. With no additional information about pricing, supply, product features, consumer demand, or launch timing, I keep the estimate near that level at 53%. The main upside risks are a compelling upgrade cycle and strong initial availability; the main downside risks are weak innovation, higher prices, economic softness, or supply constraints.

Starting from the previous forecast of 52%, I see no new concrete evidence that would justify a large update. The base rate is roughly even: newer iPhone models often sell at least as well as their predecessors because of upgrades, marketing, and installed-base growth, but mature smartphone markets and potentially strong iPhone 17 demand make a clear increase uncertain. I apply a small upward adjustment to 53% for the modest tendency of later models to benefit from a larger upgrade base, while retaining substantial uncertainty about pricing, product features, supply, and overall market conditions.

I estimate a 52% probability, assuming “sales” means comparable global unit sales during the first month after each model’s release and that iPhone 18 has a broadly comparable launch lineup and supply availability.

Prior: For a mature smartphone product, the base rate for a new iPhone generation outselling the immediately preceding generation in its launch month is roughly even—about 50%. New models benefit from upgrades, marketing, and accumulated demand, but year-over-year unit growth is not guaranteed.

Evidence increasing the probability: iPhone 18 would be the newer model and could benefit from normal replacement demand, stronger features, and possible AI-related upgrades. Apple’s brand and ecosystem also support relatively resilient launch demand. I assign this a modest positive likelihood effect, raising the odds somewhat above even.

Evidence decreasing the probability: iPhone 17 may itself have a strong launch, and the smartphone market is mature. Higher prices, supply constraints, economic weakness, or only incremental improvements could prevent iPhone 18 from exceeding its predecessor. Rumored staggered release timing for different iPhone 18 models also creates comparability and availability risk.

Combining these factors produces a posterior probability of approximately 52%. The estimate would rise materially if iPhone 18 is reported to have strong preorders, a major design or AI upgrade, and ample launch supply; it would fall if Apple staggers the lineup or iPhone 17 has unusually strong first-month demand.