Which of the following best describes the effect of overfitting on model performance?

The statement that a model excels in training data but fails in new data captures the essence of overfitting. When a model is overfit, it means that it has learned not just the underlying patterns in the training dataset but also the noise and specific details that do not generalize to unseen data. As a result, while it performs very well (or excellently) on the training dataset, it struggles to make accurate predictions when exposed to new data. This lack of generalization undermines the model's effectiveness in real-world applications, where the data may differ from that used during training.

The other choices either describe different phenomena or do not accurately depict the situation of overfitting. For instance, a model that performs equally well in both training and new data indicates good generalization rather than overfitting. Similarly, a model that is simplified for broader applicability or one that can easily adapt to changes in data suggests a desirable capability rather than the detrimental effect of overfitting.